of Independence; or if Mr. Pitt had carried his intended measures with respect to the Roman-catholics at the time of the Irish Union. So, in examining any complex case of causation, we may consider what would have happened if some circumstance which was present had been absent. For example, if the establishment of a commonwealth in England is attributed to the execution of Charles I., we may ask whether it might not have been equally established, if the king had been kept in prison or driven into exile. So, again, if a low state of prices is attributed to a combination of various circumstances, we may try to determine their relative influence, by supposing any one to be absent, and computing the probable result which the remaining elements would have produced.

These hypotheses relate to past facts, and are made for the purpose of determining real causation. All hypotheses which are made for practical purposes relate to the future, and therefore they do not attempt to determine actual facts. They are, however, framed with reference to a special case, and are equally concerned in the determination of a true sequence of events. Whenever a person is about to take any political step, or to judge of any political measure proposed by others, he frames to himself certain hypotheses as to the probable consequences which would arise from such measure, in different contingencies. Not only, however, are the consequences which he thus anticipates hypothetical, but the measure itself which is in question is of the same nature. Every project of a law, every legislative draft or bill, every proposal for a treaty, every plan of finance, is a hypothesis. It is an assumption made for the convenience of reasoning, the probable consequences of which are then examined, upon the supposition that it is to be carried into effect. The manner in which this conjectural inquiry is conducted, and the hypothetical effects of a hypothetical scheme are deduced, has been traced in a former chapter. (3) The essence of such a hypothetical inquiry is, that it is circumscribed within definite

(3) Above, ch. xiii. § 2.

bounds; that it is a practical problem, involving the application of a given plan to a specific community, under determinate circumstances.

A practical hypothesis of this sort is analogous to a tale of fiction, which is a hypothetical structure, built upon a conventional hypothetical foundation. The groundwork of the fiction may be assumed according to the taste or choice of the authorit may be placed in Paradise or in Hell; in Europe or in Asia; among the Greek deities, or in the hall of Odin; in fairy-land; in the region of the gnomes and sylphs; amidst magicians and sorcerers; in Lilliput or Brobdignag; in a desert island or a crowded city; among savage tribes, in the middle ages, or in a modern civilized society. When, however, the choice has once been made-when the conventional circumstances in which the actors of the piece are to move have once been established between the author and the reader, the compact must be adhered to, and the fundamental hypothesis of the fiction must not be violated. The author of a fictitious composition is thus analogous to the inquirer who works out the imaginary consequences of a legislative proposal: in either case, there is a hypothetical substratum, upon which a certain hypothetical superstructure is raised.

Lastly, hypotheses may be made in politics, which have reference to no special circumstances either of the past or of the future, but which are quite general in their nature, and are independent of any local or temporary circumscription. Such, for example, are all theories of an ideal state, or of any model or pattern system of political institutions. An ideal state is a supposed arrangement of a political society in a certain form, the arrangement being such, that any community might be conceived as capable of being cast into that mould. All theories of the best possible institutions are, in fact, hypothetical means, conducing to certain ends, but limited by no special conditions of time, place, or person.(*)

All theories of general causation admit of being stated, with

(4) Below, ch. xxii. § 3.

reference to the future, in a hypothetical form. In truth, this is the proper mode of expressing such a law of causation. Instead of predicting absolutely that A will produce B, it is sufficient to say, that whenever A happens B will follow. Thus, the sentence of Burke, in his Thoughts on the French Revolution'Kings will be tyrants from policy, when subjects are rebels on principle'—is the hypothetical assertion of a general law of causation, namely, that if subjects systematically resist legal power, rulers will exercise that power in a harsh manner, upon grounds of expediency. In like manner, it may be affirmed that, if certain maxims of international conduct were observed, and if independent nations abstained from certain acts towards each other, wars would cease, and a perpetual peace would prevail. Perpetual peace is a hypothetical state of things, which might exist if the conditions necessary for its existence were fulfilled; and those conditions are possible, though improbable.

§ 2 It is of the essence of hypotheses, whatever may be the nature of the subject, that they are to a certain extent arbitrary, and that they may be varied at pleasure; that different hypotheses may be started by different persons, or that one hypothesis may be substituted for another. (5) Thus, in arguing upon a case of circumstantial evidence, where the problem is to find a common cause which will explain all the appearances, rival hypotheses may be propounded. In the trial of a prisoner, for example, the hypothesis advanced on the part of the prosecution is, that his commission of the crime charged explains all the facts; whereas, his defence rests on the alleged inconsistency of some of the appearances with this supposition. Historical controversies upon the effect of evidence likewise, for the most part, consist in the attack and defence of rival hypotheses. Again, in hypotheses made for practical purposes, the same process is of perpetual occurrence: all counter-proposi

(5) When a person in argument varies the hypothesis upon which his reasoning rests without notice, he is said to shift his ground.' Arguers often shift their ground unfairly in practical discussion, in order to obviate objections made to their hypothetical propositions.

tions in affairs of business, whether public or private, are of this nature. The proposal of an amendment to a legislative draft— of a counter-project to a treaty-are merely substituted hypotheses. And as each new basis of this kind is established, a new superstructure of hypothetical arguments, showing its probable consequences in the event of its being adopted, is erected upon it.

§ 3 Uncertainty is the origin of hypotheses. If we know, we do not suppose. When a hypothesis has been proved, we It is only a hypothesis while

no longer call it a hypothesis. the argument is in a provisional state, and remains, as it were, on its trial. For example, no one would now speak of the hypothesis of the spuriousness of the Epistles of Phalaris, or the Poems of Ossian. The use of a hypothetical proposition implies doubt if not in the person who utters it, yet in the person to whom it is addressed. When the jesuit editors of Newton spoke of his assuming the hypothesis of the motion of the earth in his third book, a hypothesis which they were forced as commentators to concede, contrary to the decrees of the popes, which declared the earth to be stationary, they certainly implied that their readers would treat the hypothesis as doubtful, or even false, whatever their own belief on the subject may have been.

Hypotheses are made for the purpose of assisting the mind in reasoning, when we are uncertain as to our conclusion. We try different hypotheses in a case of circumstantial evidence, when we are in doubt as to the cause of patent effects. We try different hypotheses with reference to the future, when we are in doubt as to alternative courses of action. Again: hypotheses may be made without any specific aim, either for the past or the future, for the purpose of framing an imaginary structure of reasoning. We may assume certain premises, merely for the sake of the inferences to be drawn from them; we may establish a certain arbitrary basis, merely for the sake of the edifice to be raised upon it. Our object being merely to show the connexion of our reasoning, it is unnecessary to

assume any foundation of positive fact. It is sufficient for us to say, that if the substratum is admitted to exist, the superstructure which we build coheres with it. It is unnecessary for the contriver of an Utopia to select any real country for his legislation. All that he need say is: Given, a conceivable country, of such a territory, position, climate, fertility, population, &c., and let its political institutions be moulded after a certain supposed pattern, the people will enjoy the highest degree of happiness which civil government is capable of securing. Hence, it may be said that a proposition is false in thesi, but true in hypothesi. That is to say, a false or doubtful proposition may be admitted to be true for purposes of reasoning.

All hypotheses have this in common-that they are made for the sake of the reasoning which is founded upon them. All hypotheses, therefore, whether speculative or practical, are made for an argumentative purpose; but a practical hypothesis is argumentative, and something more. It implies that the person who makes it is in favour of the course which he supposes. When a person recommends a scheme of a law or a treaty, he not only makes a hypothesis, but he also intimates his wish that a course of conduct similar to that supposed by him should be adopted.

§ 4 Every special hypothesis ought to be within the bounds of possibility. When the argument relates to a special set of circumstances, either past or future, no benefit can arise from tracing the consequences of a hypothesis which could not have occurred previously, or which cannot occur hereafter. In discussing a case of circumstantial evidence, or in deliberating upon a practical course of conduct, our hypotheses must, in order to lead to any useful result, accord with realities. But general hypotheses are not bound by this condition. Where a hypothesis is made for a scientific purpose, imaginary data, not actual, and sometimes not even possible within the limits of our experience, may be assumed for the purpose of simplifying the problem, and of obtaining a provisional solution, which is correct for the elements it comprehends, but no further. Thus, in