Review of Bowen’s Logic. [From North American Review 99.2 (October 1865): 592-605.] ¹³
4. — A Treatise on Logic, or the Laws of Pure Thought; comprising both the Aristotelic and Hamiltonian Analyses of Logical Forms, and some Chapters of Applied Logic. By Francis Bowen, Alford Professor of Moral Philosophy in Harvard College. Cambridge: Sever and Francis. 1864. pp. xv., 450.

The publication, a few years ago, of Hamilton’s Lectures on Logic, with an Appendix containing various papers, in which his new views

on logical forms were developed, placed before the students of this science the best materials, new and old, which the genius and scholarship of the author had collected during twenty years of study and reflection, and as the fruits of a discussion which has associated his name as closely with the science of Logic as that of its great founder, Aristotle. Though we had already a foretaste of these novelties of the science in Mr. Baynes’s New Analytic of Logical Forms, and in Dr. Thomson’s excellent little treatise on the Laws of Thought, yet the great superiority of these Lectures over any work on pure Logic which had appeared in the English language gave them at once a prominent place among books of instruction, in spite of the great defects incident to a posthumous publication of writings not intended for such use, nor especially adapted to the purposes of a text-book in the instruction of classes.

To secure the excellences of these Lectures, as well as of other modern treatises, and at the same time to present their materials in a more systematic form, and within a compass convenient for a textbook, appears to have been the aim of Professor Bowen in the preparation of his treatise. In performing this important service to the study of Logic, Professor Bowen has gone over the ground of the science as it now exists in the best modern treatises. In what is by far the most important and original feature of his book, the parallel presentations of the old and new analyses of the logical elements and forms, under each of the several divisions of the subject, our author exhibits the fruits of a diligent and careful study, and we owe him much for the lucid expositions he has given of this part of the science. That skill in clear and forcible exposition which his previous writings evince is in this book turned to the best account, on subjects in which it is especially serviceable.

When we consider the unsettled state in which modern discussions of the principles and forms of Logic have left the science, and the interest which has thus been created in controverted points, at the expense of the integrity of the science, we cannot too much admire the judiciousness of the plan the author has pursued, by which he has been enabled, to include so many interesting topics under a regular and systematic development of the subject. He has, of course, assumed the position of a liberal conservatism, assenting to what the friends of the scholastic system still agree upon, and presenting in a fair and impartial manner their different views, but dissenting from views which are hostile to the science in its essential features.

The introduction of cognate terms front different authors is happily done in several instances; thus, in the use of the terms “connote” and

“denote,” which Mr. Mill revived from the usages of the schoolmen to designate the two functions of names, and in the use of the derivative terms “connotation” and “denotation” as synonymes of the “intension” or “comprehension” and the “extension” of concepts, — terms in more general use, — he has added clearness by regarding the same distinction under slightly different points of view. His treatment of the important distinction of the two quantities of concepts is, in all its applications, especially happy.

As an illustration of the unsettled state in which the science of Logic remains, we will instance the great diversity of opinion which still obtains among writers on Logic in regard to so fundamental a matter as the proper meaning of the word “inference.” While such writers as Mr. Mill would include in the meaning of the word “inference” the real force of an argument, or that which really determines belief in its conclusion, writers who follow more closely the scholastic system of Logic would confine this meaning to what they term the formal validity of an argument, or that which constitutes a truly logical procedure. But on this point there is still a want of harmony, for the advocates of the scholastic system do not agree among themselves as to what constitutes the proper distinction between a logical and a grammatical transformation of a proposition, or as to what kinds of verbal changes in a proposition should be regarded as changing its logical import or meaning. This dissension exists in respect to those inferences which are called “immediate,” in which a single proposition is supposed to be the proof of another with a different formal or logical import. The various kinds of immediate inference are all rejected by the modern or inductive school of Logic, on the ground that what, even in thought, can really determine belief in the derived proposition must be identical with what can render the original proposition credible, so that the two propositions should be regarded as identical in relation to proof. But such inferences, or some of them, are still regarded by the school logicians as having some logical value. They all agree that an interchange of subject and predicate, or a conversion, and a change of “quantity” in the terms of a proposition, are changes of meaning sufficiently important to be regarded as logical transformations, while some are disposed to regard a change of “quality” in a proposition effected by infinitating its terms, or by the double negative, as only a verbal or grammatical change. The latter appears to be the opinion of Hamilton, though not of our author. A form of immediate inference, called inference by æquipollence or infinitation, which, according to Hamilton, is of mere grammatical relevancy, is introduced with a development due to Mr. De Morgan. We cannot but regard one phase of this development

as at least a redundancy in the enumeration of simple forms of inference; tor whatever may be thought of the logical value of the double negative, a union of it with inference by simple conversion ought not to be regarded as a simple inference, even supposing it to be more than a mixture of a logical with a grammatical transformation of a proposition. Thus, to infer from “All metals are fusible,” that “All infusible substances are unmetallic,” is equivalent to “No metals are infusible” by the double negative, and hence, by simple conversion, “No infusible substances are metallic,” and lastly, by the double negative again, “All infusible substances are unmetallic.” This process may be summed up, it is true, in a single rule, which may be derived directly from logical axioms; but since it is resolvable into elementary steps, which are also given, it should not be regarded as an elementary form of inference. And we may repeat, that it should not be regarded as a proper form of inference at all, if we define logical inference, according to the inductive school, to be a relation between real grounds of belief and the proposition to be proved, formally exhibited or explicated in a reasoning. According to this definition, a reasoning indicates, but does not contain, proof, —just as a name, or its mental counterpart, a concept, indicates, but does not contain, the evidence of the coexistence of the attributes connoted by it, or of the resemblance of the things which it denotes; and just as a proposition indicates, but does not contain, the ampliative experience by which significance is added to what is already collected in the mind through the instrumentality of language.

This leads us at once to the consideration of the criticism on the mediate or syllogistic inference of the logic-books, by which modern writers have brought in question the very foundations of the science,— namely, the criticism that all syllogisms involve a petitio principii, since the truth of the premises presupposes the truth of the conclusion, and that nothing can be proved by referring a case to a rule (as is done in the syllogism), since the rule is not true unless what is supposed to follow from it is also true. Our author follows Sir William Hamilton in thinking that the analytic arrangement of the syllogism, in which, instead of the premises followed by the conclusion, the more natural order is observed, — the quæsitum or proposition in question being propounded, and the reasons made to follow, — “effectually relieves the syllogism of the imputation which has been thrown upon it for more than three centuries of being founded upon a mere petitio principii, or a begging of the question”; and this, forsooth, because “we appeal to the admitted universal truth only after we have established, what is the main point of the argument, the applicability of the truth” to the case in question. As, for example, when we argue, “Socrates is mortal, because

he is a man, and all men are mortal.” But we conceive that the more natural order of the syllogism is relieved of this imputation only in so far as it exhibits better than the synthetic order the true function of formal inference, — namely, its indication of the ground of real inference. It amounts to the same whether we virtually assume in a premise, or virtually reaffirm in a reason, the proposition we wish to prove. If the illative value of the syllogism resides in itself, then truth is nothing but verbal consistency, whatever be the order of the consistent parts. Nor does the relative importance of the premises or reasons in an argument affect this point. It is doubtless true that in most arguments the minor premise, by which a question is referred to a rule, is the one requiring emphasis, and, indeed, is the only one for which, in general, there is any need of explicit statement; nevertheless, its true function is to indicate the existence of a general rule, and, through this general rule, to indicate the existence of cases parallel to the case in question, — cases from which the rule may be justly inferred, along with the case in question. And this is the whole argument in a syllogism. It is not merely an explication of propositions, but is much more nearly what Mr. Mill describes as an interpretation of a rule to meet a case in question, though we think this description to be defective on several accounts. In the first place, it gives too great prominence, in the formal process, to the major premise, which, as our author justly thinks, is not the most emphatic part of the syllogism. In the second place, the inquiry, or the search, which leads to the discovery of an argument in answer to the question, “Why or for what reason is a given proposition true?” refers as much to the discovery of a rule through which it may be verified as to the interpretation of the rule when found, so that an argument answers, — to use Mr. Mill’s illustration, — not simply how the short-hand record of experience, the general rule, applies to the case in question, but what particular record is applicable to the case. In other words, the formal process conducts us at the same time to the discovery and the interpretation of the major premise through the minor. But the gist of the objection to the syllogism, which Mr. Mill has discussed with the greatest fairness, is, that what is expressed in the syllogism is not the whole of the process of real inference; what justifies the rule or major premise being an integrant part of the whole inference, and, indeed, the most essential part. So far from having “laboriously attempted to restrict the range, and depreciate the utility, of the syllogistic process,” as our author accuses Mr. Mill, among others, with having done, he has done much, we think, to determine its true range, and to appreciate its real utility.

It is obvious, from what has been said, that the two propositions of a so-called immediate inference should not be regarded as constituting an argument or proof proper, since neither serves to indicate more distinctly than the other the real ground on which both must be received or rejected together. It may be questioned, however, whether there are not some cases where an immediate inference will have a real validity; as when in inferring a particular proposition from a universal one in the same terms, we argue, “This cloud is composed of vapor, because all clouds are composed of vapor.” For we exclude the possible supposition that some clouds are composed of smoke or some other substance, and base the particular proposition on evidence which might not be essential to it. There is an apparent inference sometimes in these transformations, since the simpler grammatical structure indicates more clearly than the more complex one the real meaning of both, and through this the real grounds of both, so far as the meaning of a proposition can reveal the grounds of its truth or falsity. But in all true arguments a case is brought under a rule, and through the rule is brought under the evidence or authority which the rule represents; and this is oftenest effected without an explicit announcement, or even an explicit thought of the rule itself.

What words are to our first apprehensions of things, such are general propositions to that ampliative experience by which our knowledge of things is perfected, and as a word is assumed to stand equally for everything denoted by it, so a general proposition, framed of words, is assumed to hold equally for every case included under its signification, whether in actual or in possible experience. These assumptions give to words and propositions their formal value. The appeal that is made through them is not to the sum of actual experiences merely, but to the assumed universal validity of these experiences. Independently of these assumptions the knowledge we embody in language is ideal only, and truth is only consistency in thought and language. And this leads us to the consideration of the much-mooted questions, the grounds of induction and analogy and the origin of necessary truths. These questions are regarded by the followers of scholastic logic as extra-logical, and are treated in a supplement to the formal science under what is improperly called Applied Logic. As affecting real inference, they are properly regarded by the modern inductive school of logic as integrant parts of the science itself. “Why is a single instance, in some cases, sufficient for a complete induction, while in others myriads of concurring instances, without a single exception known or presumed, go such a very little way towards establishing an universal proposition?” asks Mr. Mill; and he adds, that “Whoever can answer this question

knows more of the philosophy of logic than the wisest of the ancients, and has solved the great problem of induction.” Overlooking this important question, or rather answering it after the manner of metaphysicians, by stating the facts to be explained in language which implies that they are ultimate and inexplicable, the school logicians, and among them our author, content themselves with dividing all universal propositions into two classes, — the so-called necessary truths, and the contingent or empirical truths. The former are those which require but one observation for their induction, and no experience at all, according to these writers, for their verification. With the latter or contingent truths logic proper, they say, is not concerned. These belong to the matter, and not to the form of thought; and this is also true, they admit, of the greater number of necessary truths; yet, as these are referred, not to vulgar experience, but to a special power of the mind, assumed for the purpose,—namely, the reason, nous, or locus principiorum,—they are of sufficient dignity to claim attention from philosophers. Such philosophers lump together all the degrees of certainty in experimental science, and patronizingly lend an a priori principle to eke out any deficiency in that demonstrative certainty on which they take their stand. To the vulgar empiricist there is no greater certainty than what the sum total of experience, inductive and ratiocinative, can afford; but as there are some truths in science so elementary and so incessantly presented that their contradictories cannot be represented in imagination or conceived in the understanding as possibly true, the empiricist is constrained to rise occasionally to the heights of certainty, whence, according to the logicians, all truth is derived.

Instead of dividing universal truths into the two great classes we have mentioned, a more fundamental analysis would, we conceive, establish many grades of certainty, according to the character of the evidence and the range of our experience. The determination of what we may call the inductive weights of experiments and observations — to borrow a technical term from the mathematical theory of probabilities — depends on a preliminary induction which in fact has already been performed, perhaps involuntarily and unconsciously, by every experienced mind. These inductive weights have in common the one universal presumption, that the course of nature is uniform, which is more formally explicated in what we may call, by way of distinction, the phenomenal Law of Causation, in order to discriminate it from the ontological Principle of Causality, which is quite a different matter. The presumption expressed in the Law of Causation, that all precisely similar antecedents must be followed by precisely similar consequents, should not be confounded with the Principle that nothing can be conceived

to begin to exist, or as having a really new and independent existence. The Principle of Causality is probably nothing else than a succinct but indirect statement of the history and character of all knowledge, or of the fact that everything is known along with, or as proceeding out of, something else, which in a vague and general sense is called its cause; so that we have no experience of, and no ground for representing, anything without a cause. But the character of this cause need not necessarily be that defined in the Law of Causation. And indeed this law does not possess the universal credibility or the character of necessity in thought, which belongs to the Principle of Causality. Whatever certainty the law does possess is regarded by our author as derived from this principle. The law doubtless implies the principle, but its characteristic significance, the uniformity of nature and experience, can hardly be regarded as certified by a necessary principle, unless we are prepared to admit that the law itself is also a necessary truth. The author says, “It is only necessary to show that the Law [Principle] of Causality is readily and naturally explicated into the maxim that nature’s course is uniform, so that the absolute and imperative conviction which belongs to the former as an a priori cognition of the human mind is transferred, by an easy association of ideas, to the latter, though not logically belonging to it.” The author is here obviously steering between Scylla and Charybdis. In shunning empiricism he fears to fall into fatalism. It is a new logic to us, which can explicate a maxim out of a necessary principle to which it does not logically belong; or can retain an imperative conviction in the reason and withhold it from the consequent, or transfer it only “by an easy association of ideas.” If the author is attempting to explain rather than justify the belief in the uniformity of nature, we have no grave objections, except to the phraseology of his argument, though we see no necessity for appealing to an a priori principle to account for belief in an empirical fact. Either the Principle of Causality does, or does not, prove the Law of Causation. If it does, our author is swallowed up in the yawning gulf of fatalism, and he has proved the “astounding theory,” which he quotes, in his chapter of Fallacies, from Mr. Mill, the confutation of which, he says, “is the business of the metaphysician or the theologian.” The following is the conclusion of the obnoxious passage from Mill: “The state of the whole universe at any instant we believe to be the consequent of its state at the previous instant; insomuch, that one who knew all the agents which exist at the present moment, their collocation in space, and their properties, —in other words, the laws of their agency, — could predict the whole subsequent history of the universe, at least unless some new volition of a power capable of controlling the universe should supervene.” ¹⁴ But if the Principle of Causality does not prove the Law of Causation, then the uniformity of nature must be regarded as a generalization from experience, and our author is lost in the equally obnoxious doctrine of empiricism. Independently of inductive evidence, however, this law has a formal or regulative value like that of words and general propositions. If not a universal fact, it is at least true of all we can properly learn about facts. All else is chaos. It is true as far as we know. All the evidence is for it, and of the evidence against it we are ignorant. The empiricist acknowledges his helplessness against arbitrary doubt, except in those inveterate cases where habit or primitive association in ideas is too strong to allow it. In these cases he is as resolute as the dogmatist, but his strength is the force of inertia.

Too little regard is paid, in treating of Induction and Analogy in logic-books, to what we have denominated the inductive weight of an experiment or observation. This constitutes a limitation to the general presumption of uniformity in Nature, and is of an a priori character relatively to the particular matter under investigation, but is derivable by induction and ratiocination from our experience of our liabilities to error in the given kind of matter. A single determination of the sum of two numbers by counting them together, will be regarded by any person experienced in numbers as a sufficient induction of the universality of the fact which such an experiment discloses; especially if the result be verified by any one of the many deductive processes by which the same result might be inferred. Such is the inductive weight of experiments in numbers. A physical or chemical experiment, performed with the precautions which experience has discovered, has nearly as great an inductive weight. The inductive weight in other branches of science, and in common observation, is rarely defined or definable, except in a rough way; but it is estimated and applied by all experienced observers with the same kind of subtle, unconscious sagacity which determines the bestowal of names in the formation of language.

One turns naturally to the chapter on Fallacies in the logic-books for the raciest portions of their rational festivities. Here the author descends from the lofty pure forms of thought and the contemplation of second intentions, to deal with embodied materialized forms of thought, the time-honored tricks of the dialectic art, or with the novelties of recent heresies. What the opinions of the Calvinist ministers were to the Port-Royalist logicians, such appear to be the heresies of modern theories of Natural History to our author, who essays to entrap his antagonists within the lines of pure thought, and in the very citadel of demonstration. He has managed somehow (we do not clearly understand how) to bring Mr. Darwin’s theory of “The Origin of

Species by Natural Selection” under “the Fallacy of the Composite and Divisive Sense,” but in the process a singular definition occurs: “We are often misled by the use of the word tendency. We rightly say that a given result tends to happen only when there is more than an even chance of its occurrence; if there is less than an even chance, it tends not to happen.” The application of this definition to the doctrine of the derivationists, that there is a tendency to variation in the specific characters of organic races, is obvious and fatal, because the cases of variation are greatly in the minority. But the destructive power of the definition does not stop here. The conviction we have entertained from our youth, that stones tend to roll down hill, and that the loose materials of the earth’s surface tend by every practicable route to approach nearer to the centre of the earth, must be materially modified if the definition we have quoted always holds. For the occasions of stones rolling down hill, or the occurrence of land-slides, are rather infrequent and exceptional phenomena, and cannot, therefore, according to our author, tend to happen. The existence of permanent conditions, which, if not contravened by other causes, will be constantly followed by a given result, constitutes what, in accordance with scientific usage, we should denominate a tendency, Archbishop Whately says, in treating of Ambiguous Terms, “By a tendency towards a certain result is sometimes meant, the existence of a cause which, if operating unimpeded, would produce that result.” The tendency, for example, of individual characters in animals and plants to become specific characters by inheritance. “But sometimes, again, a tendency towards a certain result is understood to mean, the existence of such a state of things that the result may be expected to take place.” The tendency, for example, in the opponents of the theory of derivation to misunderstand it. ¹⁵ and as the causes which may hinder a tendency from showing itself except in rare and occasional instances do not thereby destroy the tendency, so, on the other hand, the hypothesis of a tendency from the occasional occurrence of a phenomenon is not invalidated by the mere infrequency of the phenomenon. The real ground of induction in such cases is too subtle to be discussed under the coarse criteria of scholastic logic. And this is also true of another doctrine in natural science, which our author also burdens with a fallacy; namely, the fallacy of ignoratio elenchi, or “answering to the wrong point.” The uniformitarian school of geologists are guilty of this fallacy, the author thinks, when “they argue that the geological phenomena now visible, many of which are of stupendous magnitude, can be accounted for by the ordinary working of physical causes now in operation, if we only assign a sufficient lapse of time for the cumulation of their results,” etc. “Their ignoratio elenchi consists,” he adds, “in multiplying proofs that slow-working causes might have effected all these stupendous results, and then jumping at the conclusion that these causes did so produce them. They propound this dilemma: Accept this solution of the problem or propose a better one. We may logically decline to do either.” “An Elenchus” the author defines as “a syllogism which will confute the argument of your opponent, and ignoratio elenchi is ignorance of what will confute him, — ignorance of the fact that your conclusion, even if it were established, would not contradict his conclusion,” — and if the uniformitarians were really bent on confuting our author’s assumed logical indifference to geological theories, they might be found guilty of the fallacy. But in this the author himself ignores the point in question. The uniformitarians really adduce their arguments in confutation of the counter-arguments of those who believe that geological phenomena might have been produced by the purely hypothetical causes called Cataclysms, and then jump at the conclusion that such causes did produce them. Accept the cataclysm or propose a better solution, say the orthodox geologists, and the uniformitarians, instead of declining to do either, propose an hypothesis which more nearly accords with all that is really known of natural forces, but which demands an immense quantity of that very abundant element, Time, and as Red-Jacket said of those who complain that they haven’t time enough, “they have all there is.”

Mathematicians will be somewhat surprised to learn from the author’s chapter on the sources of evidence, etc., that “what is called ‘the Method of Least Squares ’ has been adopted as a mode of finding the most probable result, since it was ascertained that the arithmetical mean is not the best mean of a number of observed quantities [!]. This Method proceeds upon the assumption that all errors are not equally probable, but that small errors are more probable than large ones.” Now in all the treatises on this Method, it is shown to involve the principle of the rule of the arithmetical mean. Indeed, this Method is only an analytical device for computing the most probable values of such unknown quantities as are indirectly determined through the observation of other quantities on which they depend, and the most probable values in such cases — as in the cases to which the rule of the arithmetical mean is more directly applicable — are those which render the sum of the estimated errors of the observed quantities (the algebraic sum, of course) equal to zero. Or, what comes to the same, the sum of the squares of the estimated errors is required to be a minimum. This Method also gives certain conventional rules for estimating the degrees of probable accuracy in results so obtained. Our author has doubtless confounded the rule of the arithmetical mean with the method of the least absolute sum of the errors, which was used by Laplace,

and which the Method of Least Squares has superseded, with the greatest advantage to Practical Astronomy and Geodesy. But it is not easy to see why Applied Logic should take more notice of Least Squares than of the Logarithm or the Arabic system of numerals. We may add, that the rule of the arithmetical mean does not presuppose that the errors in observed quantities are all equally probable, but that any two of these quantities, considered by themselves, have an equal a priori probability of error. Considered in conjunction with the others, this probability is modified, and those observed quantities which differ least from the arithmetical mean of the whole set are regarded as the most correct, or as having probably the least error.

Among the many matters of thought through which our author has added an individual interest to his work is a doctrine almost exclusively his own, and already promulgated in his previously published writings, — namely, his doctrine of the mental constitution of the higher brute animals. Its logical interest is brought out in the Psychological Introduction of this work to illustrate the fundamental characteristics of Thought proper, the elementary operations of the Understanding, and the value of Language to that kind of knowledge it enables men to attain through its instrumentality. And the illustration is very clear and apt, provided the doctrine be admitted. Brutes not only have nothing equivalent to language proper, but they do not, according to our author, have even the elements of understanding; not a ray of true intelligence visits their darkened minds, — if mind that be which can perceive without abstracting, know without comparing, and effect what is tantamount to inference without generalizing. This theory, admitting many of the effects of abstract knowledge, denies all their known causes, and does so, we suppose, on the ground that brutes have a very imperfect comprehension of signs as such, and no proper command of them at all, and perhaps, also, because it is commendable to establish as broad a line of demarcation as possible between our intelligence and theirs. On the little logical capacity of these poor creatures our author says: “As they have only Intuitions, the only names which they can apply or understand are Proper Names, — the appellations of this or that particular, thing. These they can understand. A dog can easily be taught to know the name of his master, even when pronounced by another person. They can even be taught to know the names of particular places and buildings, so that they can understand and obey when they are told to go to the barn, the river, or the house. But it is always the particular barn, or other object, with which they have been taught to associate this sound or significant gesture as its proper name.” It would be interesting to inquire, in this connection, what happens when

one says “rat” to a terrier, or addresses the various words of command, “out,” “down,” “whoa,” and the like, to dogs and horses. Does the terrier think of the last particular victim of his sport, or the horse of his last particular act of stopping? And if so, why is the dog’s fancy apparently so filled with visions of the chase, and why does the horse stop again? That words and other signs have a generic power in all intelligences, though not always, in strict propriety, a generic significance, is a conclusion, we think, warranted by all the facts and analogies which bear upon the question. Significance is the proper attribute only of a sign cognized as such, or brought prominently under attention in its capacity as a sign, and not merely acting to direct the attention to what it suggests or reveals. The cognition of a sign, as such, involves a reversal of the natural order of association in mental acts. In our intuitions of sense, the sensations and impressions, which are the real signs of what they suggest, or direct the attention upon, are not cognized in themselves, — are not consciously cognized at all, in so far as they are significant, — but are as it were lost in the brighter light of that which they imply or reveal. The mind is unconscious of light while occupied with vision. To reverse this primitive order, to bring into equal or greater prominence in the attention that which directs attention to an object of thought, is to cognize a sign as such. But in this there is nothing added to the primitive powers of the understanding; there is rather an addition to the power of the attention. Comparison, abstraction, generalization, and even inference, depend on those fundamental laws of association common to all intelligences, through which resemblances and differences are cognized. Such acts must be relatively very imperfect when not “fixed and ratified by signs”; still the powers of understanding do not depend on language itself, but on the laws of mental association. Language is an efficient instrument of these powers, and that faculty of attention which renders it an available instrument is probably characteristic of the human mind, at least in the degree and perfection of its development.

To allow, then, that brutes can apply or understand proper names, while all cognizance of the generic power of names is denied them, is hardly a logical procedure; for Denomination is more essentially an act of clear and definite thought than Abstraction. The definition which Sir William Hamilton gives of the primum cognitum seems to us an apt description of the probable application of names in the brute intelligence. Such applications are probably “neither precisely general nor determinately individual, but vague and obscure.” A name may be applicable to all resembling objects, but will be applied on every occasion to some particular one, and may never rise to that indifferent

application to each and every one of the resembling objects which would constitute it the name of a class. For this reason it can hardly be regarded as a true name at all, even though it be applicable, like a proper name, to only one object; for in this case the sound of the word is associated with the single object in no other way than that which determines all other mental associations. But the same laws which would determine such an association would also associate the representations of resembling objects, and would direct the attention more or less definitely to their points of resemblance, and thus store the memory with generalized pictures of experience, from which would spontaneously flow such simple inferences as the actions of the higher brute animals seem, at least, to imply. And all this could take place without the instrumentality of language, or any distinct consciousness of signs, or of their significance as such.