I.

Let us first of all consider the characteristics of positive science. It

has for its object the conversion of facts into laws, or in other words

the resolving the composite into the simple, the particular into the

universal, the contingent into the necessary. But let us observe with

our author that we are only dealing here with a relative simplicity, a

partial universality, a conditional necessity. None of these characters

present themselves in a really absolute manner. The simple is invariably

composed of several terms; the universal only applies itself to a

certain class of phenomena; the necessary is so only with relation to

the consequences of a law, but the law itself always remains contingent.

Thus, no positive science can ever attain to the absolute. It is the

same with methods. These methods are induction and deduction. Now,

however precise these processes be, however marvellous the sequence and

interdependence of the propositions they discover and demonstrate, their

data are never more than particular and contingent facts; consequences,

then, can only be proportioned to those data. Hence it is certain that

the positive sciences cannot go beyond a relative universality or

necessity. It may seem as though we ought to make an exception in favour

of mathematics. But by a subtle discussion which it would be difficult

to give summarily, the author shows that they too come under the same

law, whence it follows that the domain of positive science properly

so-called is contained within the relative.

From this consideration there has sprung up in our day a philosophy that

reduces all sciences without exception to the knowledge of relation, and

by so doing has declared all metaphysics impossible: and this philosophy

is called Positivism. "Any proposition," says Auguste Comte, "which is

not finally reducible to the simple enunciation of a particular or

general fact, is incapable of holding a real or intelligible meaning."

"There is nothing absolute," says the same philosopher, "if it be not

this very proposition that there is nothing absolute." As to the proof

of this proposition, it lies, according to the school in question, in

the celebrated law which reduces all progress of the human mind in all

orders of research to three phases: the theological phase, in which

facts are explained by causes and supernatural agents; the metaphysical,

in which they are explained by abstract and ontological entities; and,

finally, the positive, in which phenomena are verified by experience and

referred to their laws--that is to say, to constant and always

verifiable relations of coincidence and succession.

Our author, having expounded this doctrine with much precision, proceeds

to criticize it with equal sagacity. He points out what is illusory in

this law of the three states; shows that it confuses metaphysic with

scholasticism; and proves, finally, that, in aiming at merging mind in

knowledge, and subordinating, as he says, the subjective to the

objective, Positivism does not understand what it is speaking of, since

all knowledge is ultimately referable to facts of consciousness--that is

to say, to something subjective, which is in effect, as Descartes has

pointed out, the only order of absolutely certain truths. Besides which,

let positive science, or rather the positive philosophy, in the name of

positive facts, proscribe metaphysic as it will, is it not evident that

the fundamental conceptions of all science--number, atom, force, matter,

cause, law--are metaphysical conceptions? Is it not evident that all

science whatever is impossible without a certain number of principles or

notions,--in a word, of intellectual laws, which even govern experience

itself? As yet the positive school has not answered the learned

demonstration of Kant on the necessity of the _à priori_ principle, or

rather it has ignored it. It has made no addition to that old

empiricism which the school of Leibnitz and of Kant had refuted.

But since the Positivism of Auguste Comte, too little versed in

metaphysical knowledge to discuss it authoritatively, there have arisen

two important schools, the one of association, the other of evolution.

The former has endeavoured to base experience on an experimental and

positive law; the latter has generalized this law, and made of it a

particular case of a more general law embracing the whole of

Nature--namely, the law of evolution.

The doctrine of association may be referred to the fundamental law that

all ideas rising simultaneously or successively in the human mind, tend

invariably to recall each other in the same order; this is what is

called association of ideas. When any two ideas have thus been

constantly associated without ever being separated (as, for instance,

form and colour), they unite indissolubly and thus become necessary

laws. Now, of all these necessary connections, the most universal is

this: no phenomenon ever appears without having been preceded by some

other phenomenon, which is always the same under the same circumstances.

This law is that of causality, which is both the supreme principle and,

at the same time, the result of all experience. To this doctrine of J.

S. Mill and Alexander Bain our author opposes the two following

objections:--1st, How does it explain the generalization? 2nd, How does

it explain the necessity of the laws of the understanding? On the first

point the English School appeals to a law that it calls the law of

_similarity_ or faculty of identifying the like in the different. But

this is indeed, strictly speaking, a fact of association? Should not

association, properly understood, be reduced to the law of

contiguity--that is to say, to the fact of our ideas only becoming

associated through relations of time? To admit the faculty of

recognizing similarity in diversity, what is this but to admit mind,

intelligence--something, in short, which is other than a simple external

association? As to the second point, can we reduce the rational

necessity that Kant and Leibnitz have laid down as the criterion of _à

priori_ principles to a pure necessity of habit--that is to say, to the

automatic expectation of the future inscribed on the past? Where is the

scientific guarantee in this hypothesis? Why should Nature bend to our

habits? "Who can assure us that we do not dream in thinking of the

future, and that the next sensation may not interrupt our dream by an

unforeseen shock?" We see how far-reaching this doubt is; it affects not

only metaphysic but science as well.

As to the philosophy of evolution, we know that, with regard to the

origin of the principles of thought, it consists in linking the

experience of present generations to that of generations past; in

substituting secular for individual experience--in a word, in filling up

by the accumulation of ages on ages the interval existing between

particular and contingent facts and the universality of principles. This

hypothesis is always at bottom no other than that of the _tabula rasa_,

only it is no longer the individual who is this _tabula rasa_, since

each one has, by heredity, received a pre-formed intelligence.

Nevertheless, under pain of contradicting the hypothesis, we are forced

to admit that there was a first subject who, prior to the action of the

object, must have been this _tabula rasa_. But here the objections of

Leibnitz reappear. What can a pure, abstract, and unmodified subject be?

And again, before any meeting of subject with object, we have to admit a

pure object having nothing subjective, just as the subject had nothing

objective. What shall we affirm of this pure object? Let us divest it if

you will of colour, heat, sound; must we not at least conceive it as

extended, as existing in time, conceive it, that is, according to the

necessary forms that are supposed to be suppressed? For to say that it

has been capable of existing without having anything in common with

these forms, and that out of this unknown and nameless condition have

arisen, by way of transformation, the notions of which we treat, were to

admit that something _can_ come out of nothing. We must therefore

acknowledge that universal notions do at least exist as germs at the

origin of evolution. It is not evolution that has created them,

evolution has only developed them, and be they ever so attenuated, they

still remain conditions without which nothing can be thought.

Such is the gist of the first part of M. Liard's book, and we have

nothing to add to it but our approbation. We can but admire the skilful

analysis with which it begins, and the vigorous discussion accompanying

that analysis. The three stages traversed by the experimental philosophy

of our days--namely, Positivism, the Associative Philosophy, and that of

Evolution--are competently and precisely summed up. The discussion is

cogent, solid, and could not be further developed without injury to the

unity of the work. No doubt it requires close attention to follow it;

but it is lucid and well sustained. Whatever the difficulty metaphysic

may encounter in constituting itself a science, and getting recognized

as such, it has been established that empiricism is not a tenable

position, since it has been found necessary to pass from positivism to

association, from association to evolution; while evolution itself still

supposed some pre-formation. One thing is certain, intelligence

invariably contains a something that does not come from without--namely,

intelligence itself.