may say that I use the term in the way indicated by its etymology: as
a standing-still on the part of thought; a clinging to old ideas after
those ideas have lost their use, and hence, like all superstitions,
have become obstructions. For I shall try to show that the doctrine
of necessity is a survival; that it holds over from an earlier and
undeveloped period of knowledge; that as a means of getting out of and
beyond that stage it had a certain value, but, having done its work,
loses its significance. Halting judgment may, indeed, at one time have
helped itself out of the slough of uncertainty, vagueness, and inadequacy
on to ground of more solid and complete fact, by the use of necessity as
a crutch; once upon the ground, the crutch makes progress slower and,
preventing the full exercise of the natural means of locomotion, tends
to paralyse science. The former support has become a burden, almost an
intolerable one.
The beginning of wisdom in the matter of necessity is, I conceive, in
realising that it is a term which has bearing or relevancy only with
reference to the development of judgment, not with reference to objective
things or events. I do not mean by this that necessity refers to the
compelling force with which we are driven to make a given affirmation:
I mean that it refers to the content of that affirmation, expressing
the degree of coherence between its constituent factors. When we say
something or other _must_ be so and so, the “must” does not indicate
anything in the nature of the fact itself, but a trait in our _judgment_
of that fact; it indicates the degree with which we have succeeded in
making a whole out of the various elements which have to be taken into
account in forming the judgment. More specifically, it indicates a
half-way stage. At one extreme we have two separate judgments, which, so
far as consciousness is concerned, have nothing to do with each other;
and at the other extreme we have one judgment into which the contents
of the two former judgments have been so thoroughly organised as to
lose all semblance of separateness. Necessity, as the middle term, is
the midwife which, from the dying isolation of judgments, delivers the
unified judgment just coming into life—it being understood that the
separateness of the original judgments is not as yet quite negated,
nor the unity of the coming judgment quite attained. The judgment of
necessity, in other words, is exactly and solely the transition in our
knowledge from unconnected judgments to a more comprehensive synthesis.
Its value is just the value of this transition; as negating the old
partial and isolated judgments—in its backward look—necessity has
meaning; in its forward look—with reference to the resulting completely
organised subject-matter—it is itself as false as the isolated judgments
which it replaces. Its value is in what it rids judgment of. When it has
succeeded, its value is nil. Like any go-between, its service consists in
rendering itself uncalled for.
All science can ultimately do is to report or describe, to completely
state, the reality. So far as we reach this standpoint regarding any fact
or group of facts, we do not say that the fact _must_ be such and such,
but simply that it _is_ such and such. There is no necessity attaching
to the fact either as whole or as parts. _Qua_ whole, the fact simply is
what it is; while the parts, instead of being necessitated either by one
another or by the whole, are the analysed factors constituting, in their
complete circuit, the whole. In stating the whole, we, as of course,
state all that enters into it; if we speak of the various elements as
_making_ the whole, it is only in the sense of making it _up_, not
of causing it. The fallacy of the necessitarian theory consists in
transforming the determinate in the sense of the wholly defined, into the
determined in the sense of something externally made to be what it is.
The whole, although first in the order of reality, is last in the order
of knowledge. The complete statement of the whole is the goal, not the
beginning of wisdom. We begin, therefore, with fragments, which are
taken for wholes; and it is only by piecing together these fragments,
and by the transformation of them involved in this combination, that we
arrive at the real fact. There comes a stage at which the recognition
of the unity begins to dawn upon us, and yet, the tradition of the many
distinct wholes survives; judgment has to combine these two contradictory
conceptions; it does so by the theory that the dawning unity is an effect
necessarily produced by the interaction of the former wholes. Only as the
consciousness of the unity grows still more is it seen that instead of
a group of independent facts, held together by “necessary” ties, there
is one reality, of which we have been apprehending various fragments in
succession and attributing to them a spurious wholeness and independence.
We learn (but only at the end) that instead of discovering and then
connecting together a number of separate realities, we have been engaged
in the progressive definition of one fact.
There are certain points upon which there is now _practical_ agreement
among all schools. What one school has got at by a logical analysis of
science, another school has arrived at by the road of a psychological
analysis of experience. What one school calls the unity of thought
and reality, another school calls the relativity of knowledge. The
metaphysical interpretation further given to these respective statements
may be quite different, but, so far as they go, they come to the same
thing: that objects, _as known_, are not independent of the process
of knowing, but are the content of our judgments. One school, indeed,
may conceive of judgment as a mere associative or habitual grouping of
sensations, the other as the correlative diversification and synthesis of
the self; but the practical outcome, that the “object” (anyway as known)
is a form of judgment, is the same. This point being held in common,
both schools must agree that _the progress of judgment is equivalent
to a change in the value of objects_—that objects as they are for us,
as known, change with the development of our judgments. If this be so,
truth, however it be metaphysically defined, must attach to late rather
than to early judgments.
I am fortunate in being able to quote from authors, who may be taken as
typical of the two schools. Says Professor Caird in his article upon
“Metaphysic,” (lately reprinted, “Essays in Philosophy and Literature,”):
“Our first consciousness of things is not an immovable
foundation upon which science may build, but rather a
hypothetical and self-contradictory starting-point of
investigation, which becomes changed and transformed as we
advance.” (“Essays,” Vol. II, p. 398.)
On the other hand, Mr. Venn writes (in the first chapter of his
“Empirical Logic”):
“Select what object we please—the most apparently simple in
itself, and the most definitely parted off from others that we
can discover—yet we shall find ourselves constrained to admit
that a considerable mental process has been passed through
before that object could be recognised as being an object, that
is, as possessing some degree of unity and as requiring to be
distinguished from other such unities.”
He goes on to illustrate by such an apparently fixed and given object
as the sun, pointing out how its unity as a persistent thing involves
a continued synthesis of elements very diverse in time and space, and
an analysis, a selection, from other elements in very close physical
juxtaposition. He goes on to raise the question whether a dog, for
example, may be said to “see” a rainbow at all, because of the complex
analysis and synthesis involved in such an object. The “mental whole” (to
use Mr. Venn’s words, the “ideal unity” as others might term it) is so
extensive and intricate that
“One might almost as reasonably expect the dog to ‘see’ the
progress of democracy in the place where he lives, of which
course of events the ultimate sensible constituents are
accessible to his observation precisely as they are to ours.”
As Mr. Venn is not discussing just the same point which I have raised,
he does not refer to the partial and tentative character of our first
judgments—our first objects. It is clear enough, however, that there
will be all degrees between total failure to analyse and combine (as,
say, in the case of the dog and rainbow) and fairly adequate grouping.
The difference between the savage whose synthesis is so limited in scope
that he sets up a new sun every day and the scientific man whose object
is a unity comprehending differences through thousands of years of time
and interactions going on through millions of miles of space is a case
in point. The distinction between the respective objects is not simply
a superimposition of new qualities upon an old object, that old object
remaining the same; it is not getting new objects; it is a continual
qualitative reconstruction of the object itself. This fact, which is the
matter under consideration, is well stated by Mr. Venn, when he goes on
to say:
“The act of predication, in its two-fold aspect of affirmation
and denial, really is a process by which we are not only
enabled to add to our information _about_ objects, _but is also
the process by the continued performance of which the objects
had been originally acquired, or rather produced_” (italics are
mine).
This statement cannot be admitted at all without recognising that the
first judgments do not make the object once for all, but that the
continued process of judging is a continued process of “producing” the
object.
Of course the confused and hypothetical character of our first objects
does not force itself upon us when we are still engaged in constructing
them. On the contrary, it is only when the original subject-matter has
been overloaded with various and opposing predicates that we think of
doubting the correctness of our first judgments, of putting our first
objects under suspicion. At the start, these objects assert themselves
as the baldest and solidest of hard facts. The dogmatic and naïve
quality of the original judgment is in exact proportion to its crudeness
and inadequacy. The objects which are the content of these judgments
thus come to be identified with reality _par excellence_; they are
_facts_, however doubtful everything else. They hang on obstinately.
New judgments, instead of being regarded as better definitions of
the actual fact and hence as displacing the prior object, are tacked
on to the old as best they may be. Unless the contradiction is too
flagrant, the new predicates are set side by side with the old as simply
additional information; they do not react into the former qualities. If
the contradiction is too obvious to be overlooked the new predicate
is used, if possible, to constitute another object, independent of the
former. So the savage, having to deal with the apparently incompatible
predicates of light and darkness, makes two objects; two suns, for two
successive days. Once the Ptolemaic conception is well rooted, cycles and
epicycles, almost without end, are superadded, rather than reconstruct
the original object. Here, then, is our starting point: when qualities
arise so incompatible with the object already formed that they cannot be
referred to that object, it is easier to form a new object on their basis
than it is to doubt the correctness of the old, involving as that does
the surrender of the _object_ (the fact, seemingly) and the formation of
another object.
It is easier, I say, for there is no doubt that the reluctance of the
mind to give up an object once made lies deep in its economies. I shall
have occasion hereafter to point out the teleological character of the
notions of necessity and chance, but I wish here to call attention to the
fact that the forming of a number of distinct objects has its origin in
practical needs of our nature. The analysis and synthesis which is first
made is that of most practical importance; what is abstracted from the
complex net-work of reality is some net outcome, some result which is of
value for life. As Venn says:
“What the savage mostly wants to do is to produce something
or to avert something, not to account for a thing which has
already happened. What interests him is to know how to kill
somebody, not to know how somebody has been killed.” (P. 62 of
“Empirical Logic.”)
And again:
“What not only the savage, but also the practical man mostly
wants, is a _general_ result, say the death of his enemy. It
does not matter whether the symptoms, i. e., the qualifying
circumstances, are those attendant on poison, or a blow from a
club, or on incantation, provided the death is brought about.
But they do desire _certainty_ in respect of this general
result.” (P. 64.)
Now it is this “general result,” the net outcome for practical purposes,
which is _the_ fact, _the_ object at first. Anything else is useless
subtlety. That the man is dead—that is the fact; anything further is
at most external circumstances which happen to accompany the fact.
That the death is only a bare fraction of a fact; that the attendant
“circumstances” are as much constituent factors of the real fact as
the mere “death” itself (probably more so from the scientific point of
view)—all this is foreign to conception. We pluck the fruit, and that
fruit is the fact. Only when practical experience forces upon us the
recognition that we cannot get the fruit without heeding certain other
“conditions” do we consent to return upon our assumed object, put it
under suspicion and question whether it is really what we took it to be.
It is, we may presume, the savage who in order to get his living, has
to regulate his conduct for long periods, through changes of seasons,
in some continuous mode, who first makes the synthesis of one sun going
through a recurring cycle of changes—the year.
As time goes on, the series of independent and isolated objects passes
through a gradual change. Just as the recognition of incompatible
qualities has led to setting up of separate things, so the growing
recognition of similar qualities in these disparate objects begins to
pull them together again. Some relation between the two objects is
perceived; it is seen that neither object is just what it is in its
isolation, but owes some of its meaning to the other objects. While in
reality, (as I hope later to point out,) this “relationship” and mutual
dependence means membership in a common whole, contribution to one and
the same activity, a midway stage intervenes before this one fact,
including as parts of itself the hitherto separate objects, comes to
consciousness. The tradition of isolation is too strong to give way at
the first suggestion of community. This passage-way from isolation to
unity, denying the former but not admitting the latter, is necessity or
determinism. The wall of partition between the two separate “objects”
cannot be broken at one attack; they have to be worn away by the
attrition arising from their slow movement into one another. It is the
“necessary” influence which one exerts upon the other that finally rubs
away the separateness and leaves them revealed as elements of one unified
whole. This done, the determining influence has gone too.
The process may be symbolised as follows: _M_ is the object, the original
synthesis of the elements seen to be of practical importance; _a_, _b_,
_c_, etc., to _h_ are predicates of constantly growing incompatibility.
When the quality _i_ is discovered, it is so manifestly incompatible
with _a_ that all attempt to refer it to the same subject _M_ is
resisted. Two alternatives are now logically open. The subject-matter
_M_, as the synthesis of the qualities _a_-_h_, may be taken up; it may
be asked whether the object is really _M_ with these qualities; whether
it is not rather Σ, having instead of the predicates _a_, _b_, etc., the
qualities ρα, ρβ, with which the new quality _i_ is quite compatible. But
this process goes against the practical grain of our knowledge; it means
not only that we do not know what we thought we knew; it means that we
did not _do_ what we thought we did. Such unsettling of action is hardly
to be borne. It is easier to erect a new object _N_, to which the more
incompatible predicates are referred. Finally, it is discovered that
both _M_ and _N_ have the same predicates _r_ and _s_; that in virtue
of this community of qualities there is a certain like element even in
the qualities previously considered disparate. This mutual attraction
continues until it becomes so marked a feature of the case that there
is no alternative but to suppose that the _r_ and _s_ of one produces
these qualities in the other, and thereby influences all the qualities
of the other. This drawing together continues until we have the one
reconstructed object Σ, with the traits ρα, ρβρ, etc. It is found that
there is one somewhat comprehensive synthesis which includes within
itself the several separate objects so far produced; and it is found that
this inclusion in the larger whole reacts into the meaning of the several
constituting parts—as parts of one whole, they lose traits which they
seemed to possess in their isolation, and gain new traits, because of
their membership in the same whole.
We have now to consider, more in detail, how the intermediate idea of
necessity grows up and how it gives away upon the discovery of the one
inclusive whole. Let us continue the illustration of the killing. The
“general result,” the death of the hated enemy, is at first the fact; all
else is mere accidental circumstance. Indeed, the other circumstances
at first are hardly that; they do not attract attention, having no
importance. Not only the savage, but also the common-sense man of
to-day, I conceive, would say that any attempt to extend the definition
of the “fact” beyond the mere occurrence of the death is metaphysical
refinement; that the _fact_ is the killing, the death, and that that
“fact” remains quite the same, however it is brought about. What has
been done, in other words, is to abstract part of the real fact, part of
_this_ death, and set up the trait or universal thus abstracted as itself
_fact_, and not only as fact, but as _the_ fact, _par excellence_, with
reference to which all the factors which constitute the reality, the
concrete fact, of _this_ death, are circumstantial and “accidental.”[66]
A fragment of the whole reality, of the actual fact individualised and
specified with all kind of minute detail, having been thus hypostatised
into an object, the idea of necessity is in fair way to arise. These
deaths in general do not occur. Although the mere death of the man, his
removal from the face of the earth, is the _fact_, none the less all
_actual_ deaths have a certain amount of detail in them. The savage has
to hit his enemy with a club or spear, or perform a magic incantation,
before he can attain that all-important end of getting rid of him.
Moreover, a man with a coat of armor on will not die just the same way
as the man who is defenseless. These circumstances have to be taken into
account. Now, if the “fact” had not been so rigidly identified with
the bare practical outcome, the removal of the hated one, a coherent
interpretation of the need for these further incidents would be open. It
could be admitted that the original death was a highly complex affair,
involving a synthesis of a very large number of different factors;
furthermore, the new cases of murder could be employed to reconstruct
the original analysis-synthesis; to eliminate supposed factors which
were not relevant, and to show the presence of factors at first not
suspected. In other words, the real fact would be under constant process
of definition, of “production.” But the stiff-necked identification of
the fragment, which happened to have practical importance with the real
object, effectually prevents any such reaction and reconstruction. What
is to be done, however, with these conditions of spear, of stone, of
armor, which so obviously have something _to do_ with the real fact,
although, as it would seem, they are not the fact? They are considered
as circumstances, _accidental_, so far as death in general is concerned;
_necessary_, so far as _this_ death is concerned. That is, wanting simply
to get the net result of the removal of my enemy, so that he will no
longer blight the fair face of nature, it is accidental how I do it;
but having, after all, to kill a man of certain characteristics and
surroundings in life, having to choose time and place, etc., it becomes
necessary, _if_ I am to succeed, that I kill him in a certain way, say,
with poison, or a dynamite bomb. Thus we get our concrete, individual
fact again.
Consider, then, that tortuous path from reality to reality, _via_
a circuit of unreality, which calls the thought of necessity into
existence. We first mutilate the actual fact by selecting some portion
that appeals to our needs; we falsify, by erecting this fragment into
the whole fact. Having the rest of the fact thus left on our hands for
disposal, when we have no need of the concrete fact we consider it
accidental, merely circumstantial; but we consider it necessary whenever
we have occasion to descend from the outcome which we have abstracted
back to the real fact, in all its individuality. Necessity is a device
by which we both conceal from ourselves the unreal character of what we
have called real, and also get rid of the practical evil consequences of
hypostatising a fragment into an independent whole.
If the purely teleological character of necessity is not yet evident,
I think the following considerations will serve to bring it out. The
practical value, the fruit from the tree, we pick out and set up for the
entire fact so far as our past action is concerned. But so far as our
_future_ action is concerned, this value is a result _to be_ reached;
it is an end to be attained. Other factors, in reality all the time
bound up in the one concrete fact or individual whole, have now to be
brought in as means to get this end. Although after our desire has been
met they have been eliminated as accidental, as irrelevant, yet when the
experience is again desired their integral membership in the real fact
has to be recognised. This is done under the guise of considering them as
means which are necessary to bring about the end. Thus the idea of the
circumstances as external to the “fact” is retained, while we get all the
practical benefit of their being not external but elements of one and the
same whole. Contingent and necessary are thus the correlative aspects
of one and the same fact; conditions are accidental so far as we have
abstracted a fragment and set it up as the whole; they are necessary the
moment it is required to pass from this abstraction back to the concrete
fact. Both are teleological in character—contingency referring to the
separation of means from end, due to the fact that the end having been
already reached the means have lost their value for us; necessity being
the reference of means to an end _which has still to be got_. Necessary
means _needed_; contingency means no longer required—because already
enjoyed.
Note that the necessity of the means has reference to an end still to
be attained, and in so far itself hypothetical or contingent, while the
contingent circumstances are no longer needed precisely because they have
resulted in a definite outcome (which, accordingly, is now a fact, and,
in that sense, necessary) and we begin to see how completely necessity
and chance are bound up with each other.
Their correlation may thus be stated: _If_ we are to reach an end we
_must_ take certain means; while so far as we want an undefined end,
an end in general, conditions which accompany it are mere accidents.
Whichever way the relationship be stated, the underlying truth is that we
are dealing with only partial phases of fact, which, having been unduly
separated from each other through their erection into distinct wholes,
have now to be brought back into their real unity.