THE PROBLEM OF MORAL FREEDOM
§ 1. The metaphysical problem of free will has been historically created
by extra ethical difficulties, especially by theological
considerations in the early Christian era, and by the influence of
mechanical scientific conceptions in the modern world. § 2-3. The
analysis of our moral experience shows that true “freedom” means
teleological determination. Hence to be “free” and to “will” are
ultimately the same thing. Freedom or “self-determination” is
genuine but limited, and is capable of variations of degree. § 4.
_Determinism_ and _Indeterminism_ both arise from the false
assumption that the mechanical postulate of _causal_ determination
by antecedents is an ultimate fact. The question then arises whether
mental events are an exception to the supposed principle. § 5.
_Determinism._ The determinist arguments stated. § 6. They rest
partly upon the false assumption that mechanical determination is
the one and only principle of rational connection between facts. §
7. Partly upon fallacious theories of the actual procedure of the
mental sciences. Fallacious nature of the argument that complete
knowledge of character and circumstances would enable us to predict
human conduct. The assumed data are such as, from their own nature,
could not be known _before the event_. § 8. _Indeterminism._ The
psychical facts to which the indeterminist appeals do not warrant
his conclusion, which is, moreover, metaphysically absurd, as
involving the denial of rational connection. § 9. Both doctrines
agree in the initial error of confounding teleological unity with
causal determination.
§ 1. The problem of the meaning and reality of moral freedom is
popularly supposed to be one of the principal issues, if not _the_
principal issue, of Metaphysics as applied to the facts of human life.
Kant, as the reader will no doubt know, included freedom with
immortality and the existence of God in his list of unprovable but
indispensable “postulates” of Ethics, and the conviction is still
widespread among students of moral philosophy that ethical science
cannot begin its work without some preliminary metaphysical
justification of freedom, as a postulate at least, if not as a proved
truth. For my own part, I own I cannot rate the practical importance of
the metaphysical inquiry into human freedom so high, and am rather of
Professor Sidgwick’s opinion as to its superfluousness in strictly
ethical investigations.[200] At the same time, it is impossible to pass
over the subject without discussion, if only for the excellent
illustrations it affords of the mischief which results from the forcing
of false metaphysical theories upon Ethics, and for the confirmation it
yields of our view as to the postulatory character of the
mechanico-causal scheme of the natural sciences. In discussing freedom
from this point of view as a metaphysical issue, I would have it clearly
understood that there are two important inquiries into which I do not
intend to enter, except perhaps incidentally.
One is the _psychological_ question as to the precise elements into
which a voluntary act may be analysed for the purpose of psychological
description; the other the ethical and juridical problem as to the
limits of moral responsibility. For our present purpose both these
questions may be left on one side. We need neither ask how a voluntary
act is performed—in other words, by what set of symbols it is best
represented in Psychology—nor where in a complicated case the conditions
requisite for accountability, and therefore for freedom of action, may
be pronounced wanting. Our task is the simpler one of deciding, in the
first place, what we mean by the freedom which we all regard as morally
desirable, and next, what general view as to the nature of existence is
implied in the assertion or denial of its actuality.
That the examination of the _metaphysical_ implications of freedom is
not an indispensable preliminary to ethical study, is fortunately
sufficiently established by the actual history of the moral sciences.
The greatest achievements of Ethics, up to the present time, are
undoubtedly contained in the systems of the great Greek moralists,
Plato and Aristotle. It would not be too much to say that subsequent
ethical speculation has accomplished, in the department of Ethics
proper as distinguished from metaphysical reflection upon the
ontological problems suggested by ethical results, little more than
the development in detail of general principles already recognised and
formulated by these great observers and critics of human life. Yet the
metaphysical problem of freedom, as is well known, is entirely absent
from the Platonic-Aristotelian philosophy. With Plato, as the reader
of the _Gorgias_ and the eighth and ninth books of the _Republic_ will
be aware, freedom means just what it does to the ordinary plain man,
the power to “do what one wills,” and the only speculative interest
taken by the philosopher in the subject is that of showing that the
chief practical obstacle to the attainment of freedom arises from
infirmity and inconsistency in the will itself; that, in fact, the
unfree man is just the criminal or “tyrant” who wills the
incompatible, and, in a less degree, the “democratic” creature of
moods and impulses, who, in popular phrase, “doesn’t know what he
wants” of life.
Similarly, Aristotle, with less of spiritual insight but more attention
to matters of practical detail, discusses the ἑχούσιον, in the third
book of his _Ethics_, purely from the standpoint of an ideally perfect
jurisprudence. With him the problem is to know for what acts an ideally
perfect system of law could hold a man non-responsible, and his answer
may be said to be that a man is not responsible in case of (1) physical
compulsion, in the strict sense, where his limits are actually set in
motion by some external agent or cause; and (2) of ignorance of the
material circumstances. In both these cases there is no responsibility,
because there has been no real act, the outward movements of the man’s
limbs not corresponding to any purpose of his own. An act which does
translate into physical movement a purpose of the agent, Aristotle, like
practical morality and jurisprudence, recognises as _ipso facto_ free,
without raising any metaphysical question as to the ontological
implications of the recognition.
Historically, it appears that the metaphysical problem has been created
for us by purely non-ethical considerations. “Freedom of indifference”
was maintained in the ancient world by the Epicureans, but not on
ethical grounds. As readers of the second book of Lucretius know, they
denied the validity of the postulate of rigidly mechanical causality
simply to extricate themselves from the position into which their
arbitrary physical hypotheses had led them. If mechanical causality were
recognised as absolute in the physical world, and if, again, as Epicurus
held, the physical world was composed of atoms all falling with constant
velocities in the same direction, the system of things, as we know it,
could never have arisen. Hence, rather than give up their initial
hypothesis about the atoms, the Epicureans credited the individual atom
with a power of occasional uncaused and arbitrary deviation from its
path, as a means of bringing atoms into collision and combination. Thus
with them “freedom of indifference” was the result of _physical_
difficulties.
In the Christian Church the doctrine seems to have owed its wide—though
not universal—acceptance to equally non-ethical difficulties of a
_theological_ kind. If God “foreknew from all eternity” the
transgression of Adam and all its consequences, how could it be
compatible with His justice to punish Adam and all his posterity for
faults foreseen by Adam’s Creator?[201] The difficulty of reconciling
the divine omniscience with the divine justice was supposed to be
avoided—in truth, it was only evaded[202]—by assuming that man was
created with a “free will of indifference,” so that obedience would have
been just as easy as transgression if man had chosen to obey. In our own
time the problem has assumed a rather different complexion, owing to the
enormous developments of mechanical physical science, which began with
Galileo and Descartes. Rigid causal determination being assumed as a
first principle of physical science, the question arose whether the
assumption should not also be extended to the psychical sphere. If so
extended, it seemed to strike at the roots of moral responsibility, by
making all human acts the inevitable “consequences of circumstances over
which we have no control”; if not admitted, the rejection of the
principle of rigid causal determination has often been thought to amount
to the denial that there is any principle of rational connection in the
psychical sphere. Hence, while persons specially interested in the facts
of the moral life have frequently inclined to the more or less radical
denial of rational connection between the events of the psychical
series, others, whose special interests have lain in the direction of
the unification of knowledge, have still more commonly thought it
necessary to hold that human action is determined by antecedents in the
same sense and to the same degree as the occurrences of the purely
physical order.
It will be our object to show that these rival doctrines of
Indeterminism and Determinism, or Necessitarianism, are alike
irrational, alike incompatible with what in practice we understand as
moral freedom of action, and alike based upon the false assumption that
rigid mechanical determination is itself an actual fact, and not a mere
postulate of the special physical sciences, valid only so far as it is
useful. But before we enter upon our task, it is necessary to begin with
a statement as to the real meaning of ethical freedom itself. Until we
know what we mean by the kind of freedom we, as moral beings, desire and
think we ought to have, it will be useless to ask whether we are or are
not free.
§ 2. “Free” and “freedom” are manifestly what are called by the
logicians “privative” terms; they denote the absence of certain
restrictions. To be “free,” in whatever special sense you may use the
word, means to be free _from_ something. What, then, are the typical
limitations which, in practice, we resent as making us unfree? They seem
to be, in the main, the following:—(1) We are not free when our limbs
are actually set in motion by an external physical agency, human or
non-human. And the reason why we are then unfree is that the resulting
movements of our bodies do not express a purpose of our own. They either
express the purpose of some other being who moves our limbs as seems
good to him, or, as in the case where we are set in motion by the
“forces” of the inanimate world, express no purpose at all that is
recognisable to us as such. And in either case _we_ have expressed no
purpose of our own by our movements; they do not truly belong to _us_ at
all, and there is therefore no freedom. It is not necessary that the
result of the movement should be one which, if it had been suggested, we
should have declined to entertain as a purpose of our own. We _might_
perhaps, if left to ourselves, have done just what another man or the
system of physical forces has done for us. Still, so long as the deed,
whatever it was, was done for us and not by us, so long as it
corresponded to no actual purpose of ours, it was not a free act.
(2) Again, we are not truly free when we act in ignorance (not due to
previous free action of our own)[203] of the special circumstances. Here
there is, as there was not in the former case, a genuine act. We
actually purpose to do something, but what we purpose to do is not the
deed which results from our movements. _E.g._, if I shoot a comrade by
mistake for one of the enemy, it is true that _I_ purpose to shoot, and
so far the shooting is an act, and a free act, of my own. But I did not
purpose to shoot my comrade, and so the result, in its concreteness, is
not the expression of my purpose, and I consequently regard myself as
not fully free in doing it, and therefore not morally accountable for
it. So far our analysis coincides with that of Aristotle, previously
referred to.
(3) Again, I am not acting freely where the circumstances are not such
as to admit of the formation of purpose at all. For this reason, merely
automatic action—if there is such a thing—is not genuine action, and
therefore not free.[204] Impulsive action without reflection, again,
comes under this category. It is, of course, accompanied by feelings of
satisfaction, and if impeded gives rise to craving, and so cannot be
called simply non-purposive. But in genuinely impulsive reaction, where
the possibility of reflection is excluded, there can be little clear
awareness of the concrete character of the purpose that is being put
into execution, and hence such action is not truly free. And in
practical life, though we are certainly held morally responsible for
impulsive action, in so far as it is thought we might have modified it
by previous habitual practice of reflection or by avoiding a situation
which we had reason to think would deprive us of the power to reflect,
we are never held as fully accountable for the deed of impulse as for
the reflectively thought out and deliberately adopted purpose.[205]
Further, we feel ourselves unfree when we fail to execute our purposes,
either from sheer inability to attend to a consistent scheme of action,
or because we attend equally to purposes which are internally
incompatible. This is why the “democratic” man, whose interests are an
incoherent medley without logical unity, and the “tyrannical man,” or,
as we should now say, the “criminal type,” whose passions are constantly
at war with one another and with his judgment, are regarded by Plato as
the typically unfree beings. To be really free, in the last resort, we
must have purposes which are coherent and abiding. And it is thus no
paradox to say that unfreedom in the end means, in the main, not knowing
your own mind, while to be free is to know what you mean.
§ 3. We may now draw some important consequences from this review of the
facts upon which every valid interpretation of freedom has to be based.
(1) Freedom, as Locke said in that famous chapter “On Power” which is
still the classic discussion of the whole subject as far as English
philosophy is concerned, “belongs to the man, not to the will.” The
proper question to ask is, “Am I free?” not “Is my will free?” or “Have
I a free will?” For “freedom” and “will,” as the facts enumerated above
show, are but the negative and the positive name for the same property,
the property of acting so as to put what we first possessed as our
private purpose into execution in the world of sensible fact. I “will”
when my outward deed is thus the expression of my purpose; in the same
case, and in no other, I am “free.” Thus to “will” and to be “free” are
one and the same thing; a will which was not free would be a will which
was not the translation into sensible fact of any one’s purpose, and
thus no will at all. Thus the question, “Are we free?” might be also put
in the equivalent form, “Can we ever will anything?” and to the
question, as thus put, experience gives a ready answer. For we certainly
do conceive purposes, and we certainly, in some of our movements, do
translate those purposes in act. And therefore we may say that freedom
is undoubtedly, in the only sense in which it is desired, a fact of
immediate experience.[206]
(2) If we retain the expression “freedom to _will_” by the side of the
phrase “freedom to _act_,” it can only be in a very special sense. It is
clear that not only may my outward deed be a translation into fact of my
present purpose, but my present purpose itself, as a psychical event,
may also be a translation into fact of a former purpose. This is largely
the case with all results of deliberate self-training and discipline,
and to a less degree with all acquired habits. Thus, _e.g._, the
movements by which I write these lines are the expression of my
preconceived purpose to write the present paragraph, but that purpose
itself, as an event in my history, is similarly the expression of a
former purpose to compose a work on Metaphysics. Thus there is a real
sense in which we can agree with Leibnitz in criticising Locke’s dictum
that we are free to act, but not free to will. For the mental conception
of a purpose is itself an act, and in so far as it translates into
existing thoughts and feelings a previous purpose it may be said itself
to be “freely willed.”[207]
(3) Freedom, in actual experience, is always limited, and, moreover,
admits of the most various degrees. As to the first point, it follows
immediately from our consideration of the circumstances which make us
unfree. If to be fully free means that your outward deed is the full
expression of an inward consistent purpose, then we can see at once that
complete freedom is, for all finite beings, an infinitely distant
_ideal_. For it means (_a_) that I am not hampered in the execution of
my purpose by vacillation of interest or conflict of incompatible
interests within myself; (_b_) nor by the establishment of “habitual”
reactions so nearly mechanical as to repeat themselves out of season
unless checked by special reflection; (_c_) nor by the limits set to my
power to “act or to forbear” in the physical world by the action of my
fellows and of “brute” nature.[208] Hence only an experience which is
absolutely devoid of internal conflict and external, partly discrepant
environment, in other words, only the experience which is the infinite
whole, can be in all its detail entirely and absolutely free. From the
possibilities of internal lack of unity of purpose and external
collision with rival purpose which are inseparable from our position as
finite beings, it must follow that we are never more than partially or
relatively free.
And that the degree to which we are free varies with the nature of our
purposes and their relation to the environment, is also manifest. There
is an indefinite plurality of such degrees, ranging up from the total or
all but total absence of freedom in the case of directly constrained
motion up to the case of cordial co-operation with the other members of
a relatively self-supporting social group in the conscious and
systematic execution of an elaborate and coherent scheme of action. To
indicate the principal distinctions among such grades of freedom which
are of practical importance for law and morality is the task of
systematic Ethics, and need not be attempted by us here. We may add that
our investigation has made it apparent that true moral freedom, of
whatever degree, is no inalienable heritage into which men step by the
“accident of birth,” but—in the main and as an actual possession—a prize
which has to be won by the double discipline of self-knowledge and
self-mastery, and of social comradeship, and may be, and is, forfeited
by the neglect of the arts by which it was first gained. No doubt one
man’s inherited disposition may make the practice of self-control, or
again of social fellowship, easier to him than to another, and to this
extent we may say that we are born with a greater or lesser “capacity
for freedom,” but of its actual possession we have all to say, “with a
great price purchased I this freedom.”
(4) Finally, our examination of the facts of morality enables us to
_define_ true freedom. We are free, as we have seen, just so far as our
experience is the embodiment of coherent and permanent interest or
purpose, and freedom is, like “will,” simply an abstract expression for
the teleological unity which, in varying degrees, is an essential
feature of all experience. Hence we can at once see that freedom does
not mean “absence of rational connection” or “absence of determination,”
but does mean, as so many recent philosophers have told us, for us
finite beings, _self_-determination. I am most free when acting for the
realisation of a coherent rational purpose, not because my conduct is
“undetermined”; in other words, because there is “no telling” what I
shall do next, but because it is, at such times, most fully determined
teleologically by the character of my inner purposes or interests,—in
other words, by the constitution of my _self_. The more abiding and
logically coherent my various purposes in action, the freer I am,
because it is my _whole_ self or system of rationally connected
interests, and not the insistence of others, or some passing whim or
impulse which I may forthwith disown as no part of my “true self,” which
is getting expression in my outward deeds. And if it were possible for a
finite being to become absolutely free, as we have seen that it is not,
such a being would, in the very moment of its entire deliverance, become
also absolutely determined from within; its whole life, as manifested to
the outsider in the series of its deeds, would become the perfect and
systematic expression of a single scheme of coherent purposes.
§ 4. We see, then, that such a genuine but limited freedom as is really
implied in the existence of morality is not only compatible with, but
actually demanded by, the principles of a sound Metaphysics. From the
side of morality we meet with the demand that human beings shall be, in
part at least, creatures whose outward acts shall be the genuine
expression of individual purpose; from the side of Metaphysics we have
already learned that just this teleological unity, genuine though
imperfect, is the essential nature of every finite experience. We are
now to see how a problem in itself quite simple leads to insoluble
difficulties and to the rival absurdities of Indeterminism and
Determinism when it is perverted by an initial metaphysical blunder. The
initial mistake of both the rival theories consists simply in taking
rigid mechanical determination of events by their antecedents in accord
with the principle of Causality as an actual fact, the divergence
between them only concerning the extent of the sphere of existence for
which such determination prevails. According to the indeterminist, the
action of conscious beings forms a solitary exception to a principle of
determination which is absolutely valid for all purely physical
processes. According to the determinist, there are no exceptions to the
principle, and our confessed inability to predict the course of an
individual life or a period of history from general laws in the same way
in which we predict an eclipse or a display of leonids, is due merely to
the greater complexity of the necessary data, and the temporary
imperfections of our mathematical methods.
It should be noted that there is no substantial disagreement between the
more sober representatives of the two views as to the actual facts of
life. The indeterminist usually admits that in practice, when you know
enough of a man’s character and of the influences brought to bear upon
him, you can tell with some confidence how he will conduct himself, and
that social intercourse, education, and penal legislation would be
impossible if you could not. Similarly, the determinist admits that it
would be very rash to treat your predictions of human behaviour in
practice with absolute confidence, and that the unexpected does
frequently happen in human life. The dispute is solely about the
philosophical interpretation of facts as to which there is virtually
universal agreement. According to the determinist interpretation, _if_
you were put in possession of the knowledge of a man’s “character” and
of his “circumstances” (and it is assumed that it is theoretically
possible to have this knowledge), and had sufficient skill to grapple
with the mathematical problems involved, you could calculate his whole
behaviour in advance, from the cradle to the grave, with infallible
precision. According to the indeterminist, you could not do so, and your
failure would arise not from any theoretical impossibility of obtaining
the supposed data, but from their insufficiency. Our behaviour, he
alleges, is not exclusively determined by the interaction of “character”
and circumstances; even with the complete knowledge of both these
elements, human action is incalculable, because of our possession of a
“free will of indifference” or power to act indifferently according to
or in violation of our “character.” You can never say beforehand what a
man will do, because of this capacity for acting, under any conditions,
with equal facility in either of two alternative ways.
I propose to show briefly that the determinist is right in saying that
conduct is completely determined by “character”—if the term be
understood widely enough—and circumstances, but wrong in holding that
this makes infallible prediction possible; on the other hand, that the
indeterminist is right in denying the possibility of such prediction,
but wrong in the reason he gives for his denial. Infallible prediction
is impossible, not because of the existence of “free will of
indifference,” but because the assumed data of the prediction are such
that you could not possibly have them until _after the event_. Finally,
it will be pointed out that the two errors both arise from the same
false metaphysical theory that the causal principle is a statement of
real fact.[209]
§ 5. _Determinism._ To begin with the view of the determinist. Human
conduct, he says, must be, like other processes, unequivocally
determined by antecedents, and these antecedents must consist of (_a_)
character and (_b_) external circumstance. For (1) to deny the causal
determination of our acts by antecedents is to deny the presence of
rational connection in the psychical sphere, and thus to pronounce not
only Psychology, but all the sciences which take psychical events as
their material and attempt to discover rational connections between
them, in principle impossible. Thus the very existence of Psychology,
Ethics, and History proves the applicability of the principle of causal
determinism to “mental states.”
(2) This is still more evident if we reflect that all science consists
in the formulation of “laws” or “uniformities,” and that the formulation
of “laws” rests upon the principle that “same result follows under same
conditions”—_i.e._ upon the principle of causal determination.
(3) Further, if psychical events are not so determined, then Psychology
and the mental sciences generally are inconsistent with the general
principles of the mechanical physical sciences.
(4) And, as a matter of fact, we do all assume that psychical events are
causally determined by their antecedents. In Psychology we assume that
our choices are determined by the strength of the _motives_ between
which we choose. Hence, if you know what are the “motives” present to a
man’s choice, and the relative strength of each, the determinist thinks
the prediction of his conduct is reduced to the purely mathematical
problem of the solution of an equation or set of equations. That our
present mathematical resources will not avail for the unequivocal
solution of such equations is, on this view, a mere temporary defect
incidental to the present condition of mathematical science. In
principle the equations must be soluble, or “there is no science of
human action.”
(5) And in practical life we do all assume that it is possible to
predict with considerable confidence the effect of typical conditions
upon the aggregate of mankind, and also, when you have the requisite
data, the effect of a definite set of conditions upon an individual man.
Thus we count upon the deterrent effects of punishment, the persuasive
influence of advertisement, etc.; and again, in proportion as we really
know our friends, we believe ourselves able to answer for their conduct
in situations which have not as yet arisen. Why, then, should we suppose
it theoretically impossible, if adequate data were furnished, to
calculate the whole career of a man or a society in advance, as the
astronomer calculates the path of a planet from its elements? These are,
I think, the chief of the stock arguments by which Determinism has been
defended. (With the purely theological argument from the absoluteness of
the divine foreknowledge I have already dealt in passing, and do not
propose to refer to it again.)
§ 6. It is not difficult to see that the logical value of all these
arguments is nothing at all. They fall of themselves into two groups,
one based upon the general view that all rational connection, or at
least all such rational connection as is significant for our knowledge,
is mechanical causal sequence, the other upon an appeal to the supposed
actual practice of the mental sciences. We may deal with the first group
(arguments 1 to 3) first. It is certainly not true that causal
determination by antecedents is the only form of rational connection.
For there is manifestly another type of connection, which we have
already seen to be fundamental for the mental sciences, namely,
teleological coherence. And we have learned in our preceding books that
no truly teleological or purposive series can really be mechanically
determined by uniform causal laws of sequence, though it is often
convenient for special purposes, as in the physical sciences, to treat
such a series _as if_ it were mechanically determined. Whether this type
of procedure will be valid in the mental sciences, depends upon the
further question whether our interest in the study of mental processes
is of the kind which would be satisfied by the formulation of a number
of abstract uniformities or laws of sequence, and the neglect of all
those features of real mental life of which such laws take no account.
In the physical sciences, as we saw, this mechanical scheme was valid
only because we have an interest—that of devising general rules for
dealing with typical physical situations—which is met by neglecting all
those aspects of concrete fact which the mechanical scheme excludes. But
we also saw that the nature of our interest in psychological
investigation was predominantly (and, in the case of the study of
voluntary action, exclusively) of a different kind. Our interest in
these investigations was to obtain such a teleological representation of
psychical processes as might be made available for the appreciative
judgments of Ethics and History and their kindred studies. Thus, even
admitting the possibility of treating psychical life for some purposes,
by abstraction from its teleological character, as if it were a
mechanical sequence, the abstraction would be fatal for the purposes of
the concrete mental sciences, and is therefore inadmissible in them. A
teleological unity in which we are interested _as_ a teleological unity
cannot, without the stultification of our whole scientific procedure, be
treated in abstraction from its teleological character.
This rejoinder to the first of the determinist’s arguments is at the
same time a refutation of the second. It is true that any science which
aims exclusively at the discovery of “laws” or “uniformities” must adopt
the causal principle, and must resolutely shut its eyes to all aspects
of concrete fact which cannot be resolved into mechanical sequence of
“same result” on “same conditions.” But, as we saw in the first chapter
of this book, the characteristic task of Psychology, except in those
parts of it which appear to be mere temporary substitutes for the
Physiology of the future, is not the discovery of “_laws_ of mental
process,” but the representation in abstract and general form of the
teleological unity of processes which are the expression of subjective
interests. Psychology, then, in its most characteristic parts, is not
based upon the causal postulate of mechanical science, but on the
conception of teleological continuity.
Our answer to the determinist’s third argument is therefore that we
admit the truth of the allegation that Psychology and all the more
concrete mental sciences which make use of the symbolism of Psychology,
because essentially teleological in their view of mental process, would
be inconsistent with the mechanical postulates, if those postulates had
any claim to admission into mental science as its ruling principles. We
deny, however, that they have any such claim to recognition. Being, as
we now know that they are, mere methodological rules for the elimination
from our data of everything which is teleological, the mechanical
postulates are only legitimate in Psychology so far as Psychology
desires mechanical results. How far that is, we have learned in the
first two chapters of the present Book, and we have found that the
initiation of purposive action is not a process which Psychology can
fruitfully treat as mechanical.
§ 7. Turning now to the determinist’s allegations as to the factual
procedure of the mental sciences, we may make the following
observations:—(1) As to the argument from the psychological treatment of
“motives” as the determining antecedents of choice, we say that it is
either an empty tautology or a fallacy, according to the sense you
please to put on the much-abused term “motive.” Choice is causally
determined by the “strongest motive”; what does this mean? If the
“strongest motive” simply means the line of action we do in fact choose,
the argument amounts to the true but irrelevant observation that we
choose what we do choose, and not something else. But if “motives” are
to be regarded as antecedents causally determining choice in proportion
to their strength, as mechanical “forces” determine the path of a
particle in abstract Mechanics, we must suppose the “strength” of the
various “motives,” like the mass of an attracting body, to be previously
fixed, independent of the choice they determine. In other words, the
determinist argument requires us to hold that alternative possibilities
of action are already “motives” apart from their relation to the purpose
of the agent who has to choose between them, and moreover have, also in
independence of the purpose or “character” of the chooser, a “strength”
which is in some unintelligible way a function of—it would not be easy
to say of _what_, though it is incumbent on the determinist to know. And
this seems no better than rank nonsense. An alternative is not a
“motive” at all, except in relation to the already existing, but not
fully defined, purpose of some agent, and whether it is a “strong” or a
“weak” motive depends likewise on the character of the agent’s purpose.
The attempt to conceive of “motives” as somehow acting on a mind with an
inherent “strength” of their own, as material particles attract other
material particles proportionately to their masses, is so palpable an
absurdity, that nothing more than the candid statement of it is needed
for its complete exposure.
And (2) there is an equal absurdity inherent in the determinist view as
to the kind of prediction of conduct which is possible in concrete
cases. We have seen already in our Third Book that no infallible
prediction of the course of events in an individual case is ever
possible. Mechanical calculation and prediction we found to be possible
in the physical sciences simply because they deal with the average
character of a vast aggregate of processes which they never attempt to
follow in their concrete individual detail. And trustworthy prediction
of human conduct by the aid of “causal laws” was seen to be of the same
kind. Your uniformities might hold good, so long as they professed to be
nothing more than statistical averages got by neglecting the individual
peculiarities of the special cases composing them, but nothing but
acquaintance with individual character and purpose would justify you in
making confident predictions as to the behaviour of an individual man.
Now, when the determinist says, “if you knew a man’s character and his
circumstances you could predict his conduct with certainty,” it is not
this kind of individual acquaintance which he has in view. He means that
the “character” of an individual man could be reduced to a number of
general formulæ or “laws of mental action,” and that from these “laws,”
by simply putting them together, you could logically deduce the man’s
behaviour. To see how irrational this assumption is, we need only ask
what is meant exactly by the “character” which we suppose given as one
of the elements for our supposed calculation. If it means the sum-total
of the congenital “dispositions” with which we are born, then—apart from
the difficulty of saying precisely what you mean by such a
“disposition”—the determinist statement is not even approximately true.
For (_a_) though it may be true that a man’s behaviour in a given
situation is an expression of his “character,” yet the “character” is
not the same thing as “congenital disposition.” Disposition is the mere
raw material of the “character,” which is formed out of it by the
influence of circumstance, the educational activity of our social
circle, and deliberate self-discipline on our own part. And the
“character” thus formed is not a fixed and unvarying quantity, given
once and for all at some period in the individual’s development, and
thenceforward constant; it is itself, theoretically at least, “in the
making” throughout life, and though you may, from personal intimate
acquaintance with an individual man, feel strongly convinced that his
“character” is not likely to undergo serious changes after a certain
time of life, this conviction can never amount to more than what we
properly call “moral” certainty, and is never justified _except_ on the
strength of individual familiarity.
(_b_) This leads us to our second point. If—to suppose the practically
impossible—you did know a man’s “character” with the knowledge of
omniscience, you would clearly also know every act of his life. For his
“character” is nothing but the system of purposes and interests to which
his outward deeds give expression, and thus to know it completely would
be to know them completely too. But—and this is what the determinist
regularly overlooks—you could not possibly have this knowledge of the
man’s “character” until you were already acquainted with the whole of
his life. You could not possibly thus know “character” as a datum given
in advance, from which to calculate, with mathematical precision, the as
yet unknown future acts of the man in question, because, as we have
seen, the “character” is, in fact, not there as a given fact _before_
the acts through which it is formed. Your data could at best be no more
than a number of “dispositions” or “tendencies,” and from such data
there can be no infallible prediction, because, in the first place,
“dispositions” are not always developed into actual fixed habits; and,
in the second, your data, such as they are, are incomplete, seeing that
“dispositions” may, and often do, remain latent and escape detection
until the emergence of a situation adapted to call them out. So that,
even if it were true that complete knowledge of a man’s original stock
of “dispositions” would enable you to calculate his career from its
elements, it would still be impossible to be sure that your knowledge of
his “dispositions” _was_ complete.
Thus, if a “science of human nature” really means a power to calculate
human conduct in advance from its elements, we must admit that there is
not and can be no such science. As a fact, however, what we really mean
by a “science of human nature,” when we speak of it as possible or as
partly existent already, is something quite different. We mean either
Psychology, individual and social, which is simply an abstract symbolism
for the representation of teleological process in its general nature, or
History, which is the detection of coherent purpose in human action,
_after the event_; or, again, Ethics and Politics, which are
appreciations of such purpose by an ideal standard of _worth_. Not one
of these sciences has ever attempted the calculation of human action in
advance by general laws; such forecasts of the future as we do make,
with rational confidence, are palpably based, wherever they are of
value, on concrete experience, our own or that of others, and not upon
the principles of an imaginary mechanics of the human mind.
§ 8. _Indeterminism._ With the fallacies of the indeterminist we must
now deal more briefly. This is the more possible as Indeterminism,
though common enough in popular moralising, has never won anything like
the position of the rival doctrine as the professed creed of scientific
investigators. The essence of the indeterminist position is the denial
of the principle affirmed alike by the doctrine of self-determination
and, in an unintelligent travesty, by the determinist theory that
conduct results from the reaction of “character” upon circumstances.
Seeing that, if all human action is mechanically determined in advance
by its “antecedents,” and is thus theoretically capable of being deduced
from its “elements,” there can be no true moral freedom, and, not seeing
that the essence of true freedom is teleological as opposed to
mechanical determination, the indeterminist thinks himself compelled to
assert that human action is, in the last resort, not “determined” even
by human character. There is a “free will of indifference” inherent in
human nature, in virtue of which a man’s acts, or at least those of them
in respect of which he is morally “accountable,” are free, in the sense
of being independent of his character.
Freedom, according to this view, consists in the ability indifferently
to adopt either of two alternative courses; so long as one alternative
is closed to you (whether by your “character” or by external
circumstances makes no difference according to the indeterminist), you
are not “free” and not acting as a moral and accountable being. You are
only acting freely in following your purpose when you could equally well
follow its direct opposite. The arguments by which this doctrine is
supported, over and above the general contention that determination by
antecedents is incompatible with moral responsibility, are chiefly of
the nature of appeals to immediate feeling. Thus we are told (1) that
when we act from choice and not under compulsion we always have the
immediate feeling that we could equally well act in the opposite sense;
and (2) that it is a matter of direct experience that, in resisting
temptation, we can and do act “in the line of _greatest_ resistance,”
and that the “will” is therefore independent of determination by
“motives.”
The detailed discussion of the actuality of the alleged facts belongs,
of course, to Psychology, and I do not propose to enter into it here.
But it should be manifest that, even admitting the facts to be as the
indeterminist states them, they do not warrant the inference he bases on
them. Thus (1) it is no doubt true that I often am aware, in resolving
on a certain course of action, that I could, _if I pleased_, act
differently. But the conditional clause by its presence makes all the
difference between teleological determination and no determination at
all. It is, _e.g._, no genuine fact of experience that I am aware that I
could violate all the habits of a lifetime, practise all the crimes I
most abhor, and neglect all the interests to which I am most devoted. I
could do all this “if I pleased,” but before I could “please” I should
have to become a different man; while I am the man I am, it is a
manifest absurdity to hold that I can indifferently express in my
behaviour the purposes which constitute my individuality or their
opposites.
(2) The argument from the successful resistance of temptations is
equally fallacious. We have seen already that the determinist assumption
against which it is directed, namely, that conduct is mechanically
determined by the inherent “strength” of “motives,” is itself unmeaning.
“Motives” are, if they are anything, another name for the interests
which constitute our character, not external influences which “work
upon” that character, and thus their relative “strength” is nothing
independent of character, but a new expression for the structure of the
individual character itself. But the counter-argument of Indeterminism
is just as unmeaning. To talk of the “conquest” of temptation as the
“line of greatest resistance” is to use the very same unintelligible
mechanical analogy as the determinist uses in talking of the antecedent
“strength” of a “motive.” There are, in fact, only two possible
interpretations of the indeterminist’s contention, and neither of them
supports his conclusion. Either the “resistance” of which he speaks must
be measured by our actual success in resisting the suggestion to act,
and in that case the very fact that we do not yield to the temptation
shows that for us yielding would have been the “line of greatest
resistance”; or else “resistance” must be measured by the extent to
which the rejected alternative still persists as a psychical fact after
its rejection. Then the alleged experience simply amounts to this, that
we can and sometimes do, in obedience to training or conviction, refuse
to act upon suggestions which as psychical facts have sufficient
intensity to remain before the mind even after our refusal. And this,
interesting and suggestive as it is, seems no particular reason for
denying the teleological determination of our conduct.[210]
The real metaphysical objection to Indeterminism however, is not that it
is an unprovable and unnecessary hypothesis, but that it involves the
denial of rational connection between human actions. By declaring that
conduct is not determined by character, it virtually asserts that it is
chance which ultimately decides how we shall actually behave in a
concrete case. And chance is simply another name for the absence of
rational connection. This is illustrated, _e.g._, by the use we make of
the conception of chance in the various empirical sciences. Thus, when I
say that it is a matter of chance what card I shall draw from the pack,
what I mean is that the result depends in part upon conditions which I
do not know, and therefore cannot use as data for a conclusion in favour
of one result rather than another. I do not, of course, mean that the
result is not conditioned at all, or that, with a sufficient knowledge
of the conditions it might not have been calculated in advance, but
merely that I in particular have not this sufficient knowledge. Hence
the admission of chance in the relative sense of “conditions not at
present accurately known” does not conflict with the fundamental axiom
of all thinking, the principle that all existence is a rational unity or
scheme of some sort. In fact, since we never can know the “totality of
the conditions” of anything, it would be true to say that there is an
element of chance, in this relative sense, in all concrete actualities.
But absolute chance, such as the doctrine of an indeterminate free will
maintains, would amount to the simple absence of any rational connection
whatever between the facts which are alleged to issue from such a will.
This is why the indeterminist view leads in the end, if consistently
carried out, to the same metaphysical absurdity as the determinist. From
failure to see that rational connection, such as is presupposed when we
impute praise or blame to an agent on the score of his conduct, means
teleological determination, both the rival theories in the end deny the
rational interconnection of human acts, the one replacing it by the
fiction of a purposeless mechanical “necessity,” the other by the equal
fiction of a “blind chance.” And the two fictions are really the same
thing under different names. For the only piece of definite information
that could be extracted either from the assertion that human conduct is
mechanically determined, or that it is the result of chance, is the
conclusion that in either case it is _not_ the expression of coherent
purpose.
§ 9. It is thus obvious that Indeterminism fails, in precisely the same
way as the opposing theory, to afford any theoretical basis for moral
responsibility. True, I cannot be “responsible” for deeds which are the
outcome of a purely mechanical system of antecedents, because such
deeds, not issuing from the purposes of my self, are in no true sense
mine; but the same would be equally true of the results of an
indeterminate free will. As not owing their existence to my purpose,
those results are in no real sense “my” acts, and the choice of the name
“free will” for their unknown source only serves to disguise this
consequence without removing it. Only as issuing from my character, and
as the expression of my individual interests, can acts be ascribed to me
as “mine” and made the basis of moral approbation in censure of my
“self.”
Thus we see that the determinist and the indeterminist are led alike to
impossible results because of the common error involved in their point
of departure. Both start with the false assumption that the causal
determination of an event by its “antecedents”—which we have in our
earlier books seen to be a postulate ultimately not in accord with
reality, but permissible in so far as it permits us to obtain useful
results by treating events as if they were thus determined—is ultimately
real as a feature of concrete existence. Having thus at the outset
excluded genuine teleological determination from their conception of the
world of change, both theorists are alike debarred from the correct
understanding of those psychical processes for the comprehension of
which teleological categories are indispensable.
In the terms of theories which treat determination as purely mechanical,
the factors which manifestly are the determining conditions of conduct,
namely, character and the alternative possibilities of action,
inevitably come to be conceived of as the temporal “antecedents” of the
act which issues from them. And when once this notion of character as a
sort of pre-existing material upon which “motives” from without operate
has been framed, it matters little in principle whether you take
“character” and “motive” by themselves as the complete antecedents by
which action is determined, or add a third “antecedent” in the form of
an inexplicable arbitrary “free will.” In either case all possibility of
a truthful representation of the freedom actually implied in moral
accountability was surrendered when the “character” which expresses
itself through an act, and the “motive” which is another name for that
character as particularised by reference to circumstances, were falsely
separated in thought from each other, and then further treated as the
temporal antecedents of the act in which they are expressed. In our own
treatment of the problem of freedom we were able to escape both sides of
the dilemma, because we recognised from the first that the categories of
mechanical determination are not the expression of real fact, but
limitations artificially imposed upon facts for special purposes of a
kind which have nothing in common with the ethical and historical
appreciation of human conduct, and therefore irrelevant and misleading
when applied out of their rightful sphere.
_Consult further_:—H. Bergson, _Sur les données immédiates de la
conscience_; F. H. Bradley, _Ethical Studies_, Essay 1; W. R. B. Gibson,
“The Problem of Freedom” (in _Personal Idealism_); T. H. Green,
_Prolegomena to Ethics_, bk. i. chap. 3, bk. ii. chap. 1; W. James,
_Principles of Psychology_, vol. ii. chap. 26; _Will to Believe_ (_The
Dilemma of Determinism_); J. Locke, _Essay concerning Human
Understanding_, bk. ii. chap. 21 (on _Power_); J. Martineau, _Types of
Ethical Theory_, vol. ii. bk. i. chap. 1; J. S. Mill, _Logic_, bk. vi.
chap. 2 ff.; J. Royce, _The World and the Individual_, Second Series,
lect. 8; H. Sidgwick, _Methods of Ethics_, bk. i. chap. 5; _Lectures on
the Ethics of Green_, etc., pp. 15-29.
-----
Footnote 200:
See _Methods of Ethics_, bk. i. chap. 4, § 6 (pp. 72-76 of 5th ed.).
Footnote 201:
So Omar Khayyám—
“Oh Thou, who didst with pitfall and with gin
Beset the Road I was to wander in,
Thou wilt not with Predestined Evil round
Emmesh, and then impute my Fall to Sin.”
(FITZGERALD, ed. 4, stanza 80.)
And our own poet—
“Thou madest man in the garden; thou temptedst man, and he fell,” etc.
(For the original of the stanzas on Predestination in Fitzgerald’s
Omar, see, _e.g._, the Persian text of Whinfield, quatrains 100, 126,
197.)
Footnote 202:
Evaded, because, even granting the satisfactoriness of the solution
for the special case of Adam, there would still be the problem of
reconciling the alleged “free will” of his descendants with their
inheritance of “original sin.” The more rigid Calvinism, with its
insistence on the natural corruption of man’s heart and the
absoluteness of predestination, seems to secure logical consistency at
the expense of outraging our moral convictions. Like so many popular
theological problems, this of the conflict between God’s omniscience
and justice arises from a misconception of the issue. It is only when
the category of time is illogically applied to the _ex hypothesi_
perfect, and therefore timeless, nature of God that God’s knowledge
comes to be thought by as _fore_ knowledge _before the event_, and
thus occasions the difficulty which the “free-will” theory was
intended to remove. See on this point, Royce, _The World and the
Individual_, vol. ii. lect. 8, and compare Bradley, _Ethical Studies_,
p. 19. Of course, the case would be altered if we thought of God as
finite and imperfect, and therefore in time. But there would then be
no longer any reason for believing either in His omniscience or His
omnipotence, and so no problem would arise.
Footnote 203:
Remember that abstention from acting is itself action, just as in
Logic every significant denial is really an assertion. Hence our
proviso meets the case of wilful neglect to inform myself of the
material circumstances.
Footnote 204:
The only automatic acts of which we really _know_ the psychical
character are our own “secondarily automatic” or “habitual” acts. It
is, of course, a problem for the casuist how far any particular
reaction has become so completely automatic as to be no longer an
occasion for the imputation of merit or guilt.
Footnote 205:
For purposes of law it may often be impossible to draw the
distinction, and we may have to acquiesce in the rough-and-ready
alternative between entire accountability and complete
non-accountability. But in passing moral judgment on ourselves or
others _in foro conscientiæ_, we always recognise that accountability
is a thing of degrees. On this point see Mr. Bradley’s previously
quoted article in _Mind_ for July 1902.
Footnote 206:
It must, however, be carefully noted that will in the sense in which
it is equivalent to freedom must be taken to include what some
writers, _e.g._, Bradley, call a “standing” will—_i.e._ any series of
acts originally initiated by an idea of the resultant changes, which
is approved of by us unconditionally. In the actual execution of such
a series of acts many of the stages are habitual reactions which, as
such, are not accompanied by the “idea” of their specific result as a
determining condition of their occurrence. The sphere of moral freedom
is arbitrarily restricted when it is assumed that an _actual_ volition
is indispensable for every stage of the “free” action.
Footnote 207:
The reader should study for himself Locke’s famous chapter (_Essay_,
bk. ii. chap. 21). Locke’s treatment, hampered as it is by his
unfortunate retention of the discussion of his first edition side by
side with a somewhat modified re-statement, compares favourably for
clearness and sound sense with that of most subsequent philosophers,
notably with Kant’s unintelligible attempt to reconcile the absolute
freedom of man as “noumenon” (a fictitious quality of a fictitious
being) with his equally absolute unfreedom as “phenomenon” (another
equally palpable fiction).
For Leibnitz’s criticism of Locke, see _Nouveaux Essais_, II. xxi.,
particularly §§ 8-25. (The English translation by Langley can only be
used with extreme caution.) On the whole question the reader should
also consult Green, _Prolegomena to Ethics_, bk. ii chap. 1; Bradley,
_Ethical Studies_, Essay 1, and article in _Mind_ for July 1902; W. R.
B. Gibson, “The Problem of Freedom” (in _Personal Idealism)_.
Footnote 208:
Then, are “animals” free? I see no reason to deny that, since their
life, in as its degree, must have teleological continuity to be a life
at all, they too must possess a rudimentary degree of freedom, though
a degree not sufficient to fit them for a place as ἴσοι καὶ ὄμοιοι in
human society, and therefore, for the special purposes of human
ethical systems, negligible. Similarly, a human imbecile may possess a
degree of freedom which is important for the educator who is
interested in the “care of the feeble minded,” and yet may rightly be
treated for the different purposes of a penal code as simply unfree.
Footnote 209:
Compare with what follows, Bradley, _Ethical Studies_, Essay 1, and
the notes appended to it. For a typical statement of the determinist
case in its more sober form, see Mill, _System of Logic_, bk. vi.
chap. 2. It is harder to find a reasonable statement of the opposite
view, as most capable moral philosophers have adopted the doctrine of
self-determination. For a defence of thoroughgoing Indeterminism, see
James, _The Will to Believe_ (Essay on _The Dilemma of Determinism_).
In Professor Sidgwick’s statement of the indeterminist view (see,
_e.g._, his posthumous lecture on T. H. Green’s doctrine of freedom in
_Lectures on the Ethics of Green, Spencer, and Martineau_, pp. 15-28),
Indeterminism seems to me to be qualified to the point of being in
principle surrendered.
Footnote 210:
See the admirable discussion of this experience in Dr. Stout’s _Manual
of Psychology_,^3 bk. iv. chap. 10, § 7.
-----