GRICE E CAPOCASALE: AD PLACITVM

 AN     OUTLINE     OF     SEMATOLOGY;     OR     AN ESSAY     TOWARDS ESTABLISHING A     NEW THEORY     OP     GRAMMAR, LOGIC, AND RHETORIC,     " Perhaps if words were distinctly weighed and duly considered, they  would afibrd us another sort of Logic and Cretic, than what we have been  hitherto acquainted w4th." — Locke.     LONDON :  JOHN RTCHARDSON, ROYAL EXCHANGE.     1831.     G WOODPALL, AHQEh COUBT, •KllfWl* tTRWT, LOWDOH.      ADVERTISEMENT.     I PUT not my name to these pages, nor shall  I, beyond this notice, speak in the first per-  son singular, but assume the pomp and cir-  cumstance of the editorial "we". Why I  choose for the present to remain unknown, I  leave the reader to settle as his fancy pleases.  He is at liberty to think that, being of no  note or reputation, and fearing for my book  the fate of George Primrose's Paradoxes, I do  not place my name in the title page, because  it would inevitably make that fate more cer-  tain. Or, if he chooses, he may imagine a  better motive. He may suppose me to be  the celebrated author of ***** *, with half  the alphabet in capitals at the end of my  name ; and that I prefer an incogfiito, lest  he, my " cotirteous reader", should relax the  rigour of examination, and receive as true,  on the authority of a name, a theory that  may be false.     OUTLINE OF SEMATOLOGY.     INTRODUCTION.  In the last chapter of Locke's Essay on the  Human Understanding , there is a threefold  division of knowledge into ^uo-t*^, TrpaxriK^,  and trtjfieiaTiK'^. If we might call the whole  body of instruction wliich acquaints ua with  TO. <f>v<TtKa by the name Physicology, and  that which teaches to -irpaKTixa by the name  Practkology, — all instruction for the use of  TO <7?j^aTo, or the signs of our knowledge,  might be called Sematology *.   * Physicology, far more comprehensive than the  sense to wliich Physiology is fixed, would in this case  signify the doctrine of the nature of all things what-  ever which exist independently of the mind's concep-  tion of them, and of the human will ; which things in-  clude all whose nature we grow acquainted with by ex-  perience, and can know in no other way, and therefi>re  include the mind, and God ; since of the mind as well  as of sensible things we know the nature only by ex-  perience, and since, abstracted from Revelation, we  know the existence of a God only by experiencing His  providence, Practicology, the next division, is the  doctrine of human actions determined by the will to s  preconceived end, namely, something beneficial to in-  dividuals, or to communities, or the welfare of the  kJ     INTRODUCTION.     The signs which the mind makes use of  in order to obtain and to communicate know-  ledge, are chiefly words ; and the proper and  skilful use of words is, in different ways, the  object of, 1. Grammar, of 2. Logic, and of  3. Rhetoric. Our outline of Sematology  will therefore be comprised in three chap-  ters, corresponding with these three di-  visions.   species at large. As to Sematology, the third division,  it is the doctrine of signs, showing h ow the mind ope-  rates by their means in obtaining the knowledge com-  prehended in the other divisions. It includes Meta-  physics, when Metaphysics are properly limited to  things TB /*ETa Tct pi/fiKa, i. e. things beyond natural  things — things which exist not independently of the  mind's conception of them ; e. g. a line in the abstract,  or the notion of man generally: for these are merely  signs which the mind invents and uses to carry on a  train of reasoning independently of actual existences ;  e. g. independently of lines in concrete, or of men in-  dividually and particularly. But as to the class of  signs which the former of these instances has in view,  and which are peculiar to Mathematics, there will be  no necessity, in this treatise, to make much allusion to  them: it is to the signs indicated by the other example  that reference will chiefly be made: for these are the  great instruments of human reason, and we believe  they have never yet had their suitable doctrine.     ■■>.l ■■ ■. ■ ■ ■   ■ ■■.■•••1 : ^'. .■   h . CHAPTER I.   ON GRAMMAR.     y- ■' •*     —reveal MEPOnXlN avdf wsrwy. £[oM£E.     T ■     1. To ascertain the true principles of Gram-  mar, the method often pursued will be adopt-  ed here j namely, to imagine the progress of  speech upward as from its first invention. As  to the question, whether speech was or was  not, in the first instance, revealed to man, we  shall not meddle with it : we do not propose  to inquire how the first man came to speak ^^   ^ Beattie and Cowper, poets if not philosophers, ate  among those who insist that speech must have been  revealed. The former thus turns to ridicule the well   L   known passage in the Satires of Horace, Cvm pro-   repseruntf &c. lib. I. Sat 3* v. 99 : —  ^^ When men out of the earth of old  A dumb and beastly vermin crawled.  For acorns, first, and holes of shelter, •  They, tooth and nail, and bdter dceker,   B 2     4 ON CiSAUMAH. [CHAP. I.   but whether language is not a necessary effect  of reason, as well as its necessary instrument,  Fought fist to fist ; then with a club  Each learned hia brother brute to drub ;  Till more experienced grown, these cattle  Forged fit accoutrements for battle.  At last, (Lucretius Bays, and Creech,)  They set their wits to work on speech :  And that their thoughts might all have marks  To make them known, these learned clerks  Left ofi' the trade of cracking crowns,  And manufactured verba and nouns."   Theory of Language, Part I.  Chap 6. (in a note.)  The other poet does not, on this occasion, appear in  metre, but is equally merry.   " I ta';e it for granted that these good men are phi-  Bophically correct in their account of the origin of  language ; and if the Scripture had left us in the dark  upon that article, I should very readily adopt their  hypothesis for want of better information. I should  suppose, for instance, that man made his first effort in  speech in the way of an interjection, and that ah ! or  oh ! being uttered with wonderful gesticulation and  variety of attitude, must have left hia powers of ex-  presdon quite exhausted ; that, in a course of time, he  would invent many names for many things, but first  for the objects of his daily wants. An apple would  consequently be called an apple ; and perhaps not     SECT. 1.] ON GRAMMAR. 5   growing out of those powers originally bestow-  ed on man, and essential to their further deve-  lopment.   many years would elapse before the appellation would  receive the sanction of general use. In this case, atid  upon this supposition, seeing one in the hand of  another man, he would exclaim, with a most moving  pathos, * Oh apple !' Well and good, — ' Oh apple,** is  a very affecting speech, but in the mean time it profits  him nothing. The man that holds it, eats it, and he  goes away with ' Oh apple!** in his mouth, and nothing  better. Reflecting on his disappointment, and that  perhaps it arose from his not being more explicit, he  contrives a term to denote his idea of transfer,, or  gratuitous communication, and the next occasion that  offers of a similar kind, performs his part accordingly.  His speech now stands thus — * Oh give apple ! ** The  apple-holder perceives himself called upon to part with  his fruit, and having satisfied his own hunger, is  perhaps not unwilling to do so. But unfortunately  there is still room for a mistake, and a third person  being present, he gives the apple to him. Again dis-  appointed, and again perceiving that his language has  not all the precision that is requisite, the orator retires  to his study, and there, after much deep thinking,  conceives that the insertion of a pronoun, whose office  shall be to signify, that he not only wants the apple to  be given, but given to himself, will remedy all defects ;     6 ON GRAMMAR. [CHAP. I.   S. Now instead of taking it for granted, as  others have done who have pursued the method  proposed, that men sat down to invent the  parts of speech, because they found they had  ideas which respectively required them, we as-  sert that men have originally no such ideas as  correspond to the parts of speech. The im-  pulse of nature is, to express by some single  sound, or mixture of sounds (not divisible in-  to significant parts) whatever the mind is  conscious of; nor is there any thing in the na-  ture of our thoughts that leads to a different  procedure, till artificial language begins to be   he uses it the next opportunity, succeeds to a wonder,  obtains the apple, and, by his success, such credit to  his invention, that pronouns continue to be in great  repute ever afl^er. Now as my two syllable-mongers,  Bcattie and Bl^r, both agree that language was  originally inspired, and that the great variety of  languages we find on earth at present, took its rise from  the confusion of tongues at Babel, I am not perfectly  convinced, that there is any just occasion to invent  this very ingenious solution of a diiEculty, which  Scripture has solved already."   Letter to the Rev. Wm. Unwin, April 5, \'J8i.     SECT. 2.] ON GRAMMAR. 7   invented or imitated. Let us take, for our first  fact, the cry for food of a new-born infant: that  is an instinctive ciy, wholly unconnected, we  presume, with reason and knowledge. In pro-  portion as the knowledge grows, that the want,  when it occurs, can be supplied, the cry be-  comes rational, and may at last be said to sig-  nify, " Give me food," or more at full," I want  you to give me food." In what does the ra-  tional cry, (rational when compared with the  instinctive cry,) differ from the still more ra-  tional sentence? Notin its nieaning,but simply  thus, that the one is a sign suggested directly  by nature, and the other is a sign aijsing out  of such art, as, in its first acquirement, (we are  about to presume,) nature or necessity gradu-  ally teaches our species. Now, that the arti-  ficial sign is made up of parts, (namely the  words that compose the sentence,) and that  the natural sign is not made up of significant  parts, we affirm to be simply a consequence of  the constitution of artificial speech, and not to  follow from any thing in the nature of the com-     ON GRAMMAR.     [chap. I.     munication which the mind has to make. The  natural cry, if understood, is, for the purpose  in view, quite as good as the sentence, nor  does the sentence, as a whole, signify any thing  more. Taking the words separately, there is  indeed much more contained in the sentence  than in the cry ; namely, the knowledge of  what it is to give under other circumstances  as well as that of giving food ; — oi'Jbod un-  der other circumstances as well as that of be-  ing given to me; — of me under other circumt \  Btances as well as that of wanting food:  but all this knowledge, in this and similar  cases for which a cry might suffice, is un-  necessary, and the indivisible sign, if equally  understood for the actual purpose, is, for  this purpose, quite adequate to the artificially  compounded sign.   S. The truth is this, that every perception  by the senses, and every conception* which     • *' By Conception I mean that power of the mind,  which enables it to fonn a notion of an absent object of  perception ; or of a sensation which it has formerly     SECT. 3.] ON GRAMMAR. 9   follows from such perception, as well as every  desire, emotion, and passion arising out of  them, is individual and particular; and if lan-  guage had continued to be nothing more than  an outward indication of these its passive affec-  tions, it would have consisted of single indivi-  dual signs for single individual occasions, like  those which are originally prompted by na-  ture. But it was impossible to find a new sign  for every new occasion, and therefore an ex-  pedient was of necessity adopted; which ex-  pedient, from its rudest to its most refined ope-  ration, will be found one and the same, — an  expedient of reason, and that through which  all the improvements of reason are derived.  The expedient is nothing more than this : —  when a new expression is wanted, two or more  signs, each of which has served a particular  purpose, are put together in such a manner as  to modify each other, and thus, in their united     fclt." — Dugald Stewart : I'hilos. of the Human Mind,  Vol. I. Chap. 3.     10     ON GRAMMAR.     [chap. I.     capacity, to answer the new particular purpose  in view. In this manner, words, individually,  cease to be signs of our perceptions or con-  ceptions, and stand (individually) for what are  properly called notions', that is, for what the  mind knows ; — collectivelif, that is, in sen-  tences, they can signify any perception by the  senses, or conception arising from such per-  ception, any desire, emotion, or passion — in  short, any impression which nature would  have prompted us to signify by an indivisible  sign, if such a sign could have been found : —  but individually, (we repeat,) each word be-  longing to such sentence, or to any sentence,  is not the sign of any idea whatever which the  mind passively receives, but of an abstractiont   • Notio or notitia from «o«co, I knov. (It is a pity  we cannot trace the word to ado instead of noac.->.)  Note, Locke will be mucli more intelligible, if, in the  majority of places, we substitute " tlie knowledge of"  for what he calls " the idea of" His wide use of the  word idea has been a cause of the widest con&slon in  other writers.   t Home Tooke's doctrine is very different from     SECT. 3.] ON GRAMMAR. 11   wliich reason obtains by acts of comparison and  judgment upon its passively-received ideas.   tbis. He says (Diversions of Purley [2d edit. 1798]  Vol. I. page 51,) " That the business of the mind, as  far as regards language, extends no further than to re-  ceive impressions, that is, to have sensations or feel-  ings"; — he affirms (pa££^im) that what iscalled abstrac-  tion has no existence in the mind, but belongs to lan-  guage only, and that " the very term metapht/sic is  nonsense "' {page 399). It is hoped that what follows in  the test will prove these opinions to be erroneous.  Could the proper name John, or any word being an ar-  tificial part of speech, have been invented, if the mind  had not exerte  d its active powers upon its passively r&-  ceived ideas ? For whatever ideas of this last kind we  have of John must be ideas arising out of particular  perceptions ; and ve must irame him to our minds  standing, or sitting, or walking; talking, or silent;  dressed or undressed, with other circumstances which  imagination can vary, but cannot set aside. It is only  by comparison that we know John to be independent  of all these, and the name is the effect of this know-  ledge, not the cause of it. The abstraction is not in  the word only ; for till we know that Jolm is separate  (abstract) from whatever circumstance the perception  of him includes, how can his name exclude it ? Neither  is the terra iiietaphysic nonsense when applied to this     ON GRAMMAlt.     [chap.     The sentence " John walks " may express  what is actually perceived by the senses ;     or any other abstraction. For John separate from cir-  cumBtancea that must enter into an actual perception,  ifithe nameof anotion /iCTa^ua-ixii, i.e.outof nature, or of  which we have no example in external nature, though  it may esist in our minds, like a line in mathematics,  which is deifined as that which has length without  breadth, and which is therefore, for the same reason,  properly called a metaphysical notion, and pure  mathematics are justly considered a part of metaphysics.  It was because H. Tooke set out with these principles  thus fiindamentally erroneous, that he could not com-  plete his system when he had brought it to ail but a  close. With admirable acuteness of inquiry, he had  tracedup every part of speech till he found it, originally,  either a noun or a verb, and he then left his book im-  perfect, because he could not, on the principles he had  started with, explain the difference bet ween these : — he  promised indeed to return to the inquiry, but he never  fiiliilled his promise for the best of reasons, that there  was no pushing it further in the way he had gone ; he  must have contradicted all his early premises to have  reached a true conclusion. The whole cause of his  error seems to havebeen a too unqualified understanding  of Locke's doctrine, that the mind has no innate ideas.     SECT. 3.3     ON GRAMMAR.     but neither word, separately, can be said to  express a part of that perception, since the  perception is of John walkmg, and if we per-  ceive John separate from walking, then he is  not walking, and consequently it is another  perception ; and so if we perceive walking se-  parately from John, it must be that we per-  ceive somebody else walking, and not him.  The separate words, then, do not stand for  passively received ideas, but for abstract no-  tions ; — so far as they express what is pec- ij  ceived by the senses, they have no separate  meaning ; it is only with reference to the un-  derstanding that each has a separate meaning.  The separate meaning of the word John is a  knowledge (and therefore properly called a I  notion not an idea*) that John has existed and ]   Hence, Tooke acknowledges nothing originally but ]  the senseB, and the experience of those senses, calling ■  reason " the effect and result of those senses and that  experience." See Vol, II. page 16.   " If indeed the word idea were uniformly employed  to signify what is here meant by notion, and nothing  else, little objection could be made : such use would     14     fCHAP. r.     will exist, independently of the present per-  ception, and the separate meaning of the word  •walks, is a linowledge that another may waik as  well as John. This is not an idea of John or an  idea of walking such as the senses give, or such  as memory revives : for the senses present no  such object as John in the abstract, that is,  neither walking, nor not walking ; nor do they  furnish any such idea as that of •walking inde-  pendently of one who walks. There is then  a double force in these words, — their separate  force, which is derived from the understanding,  and their united force, by which, in this in-  stance, they signify a perception by the senses.   nearly correspond in effect though not in theory, with  the old Platonic Bcnse, and in the Platonic sense  Lord Mooboddo constantly employs it in his work on  the " Origin and Progress of Language." But as Dr.  Reid observes, ** in popular language idea signifies  the same thing as conception, apprehension. To have  KD idea of a thing is to conceive it." This sense of  the word Dugald Stewart adopts. (Philos. of the  Human Mind, Vol. L Chap. 4. Sect. 2.) Locke, as  already intimated, uses the word in all the senses it  will bear.     SECT. 4.3 ON GRAMMAR. 15   4. In otlier instances, the united significa-  tion of words may not be a perception of the  senses j but whatever may be their united  meaning, they will separately include know-  ledge not expressed by the whole sentence,  though, if the meaning of the sentence be ab-  stract, the knowledge included in the separate  words will be necessary to the knowledge ex-  pressed by the sentence. " Pride offends,"  is a sentence whose whole meaning is abstract;  but pride separately, and offends separately,  are still more abstract, and in using them to  form the sentence, we refer to knowledge be-  yond the meaning of the sentence as a whole,  namely, to pride under other circumstances  than that of offending, and to offending under  other circumstances than that of pride offend-  ing ; and here, tlie knowledge referred to  seems necessary, in order to come at the know-  ledge expressed by the sentence. " John  walks," (or, according to our English idiom,  " John is walking,") is a perception by the  senses, and does not therefore depend on a     16     01^ GRAMMAR.     [chap.     knowledge of John, and of walking in the ab-  stract ; (though to express the perception in  this way requires it;) but " Pride offends,"  does not express an individual perception, nor  would many individual perceptions of pride  offending give the knowledge which the sen-  tence expresses : we must have obser\'ed  what pride is, separately from its offending,  and we must have observed what offending is,  separately from pride offending, before we  can rationally understand, or try to make  known to others, that Pride offends. In this  DOUBLE force of words, by which they signify  at the same time the actual thought, and re-  fer to knowledge necessary perhaps to come  at it, we shall find, as we proceed, the ele-  ments, the true principles of Logic and of  Rhetoric; while in tracingthe necessity which  obliged men to signiiy in this manner even  tliose individual perceptions which nature  would have prompted them to make known  by a single sign, (if such sign could have been  found,) we shall ascertain the true principles     SECT. 5.j     ON GRAMMAR.     17     of Gkammau. The last mentioned subject  must occupy our first attention.   5. To get at the parts of speech on our hy-  pothesis, we must consider them to be evolved  from a cry or natural word. Not that this  is the present principle on which words are  invented ; for art having furnished the pattern,  we now invent upon that pattern j but our  purpose is to consider how the pattern itself  is produced by the workings of the human  mind on its first ideas. Those ideas can be  none other than the mind passively receives  through the senses ; and perhaps the first ac-  tive operation of the mind is to abstract (sepa-  rate) the subjects or exterior causes of sensa-  tion from the sensations themselves. When  we see, we find we can touch, or taste, or  smell, or hear ; and when the perception  through one of these senses is different, we  find a difference in one or more of the others.  We also recollect (conceive) our former per-  ceptions, and finding the actual sensations  not recoverable by an effort of the mind alone,     18     ON GBAWMAR.     [chap. I.     we recognize the separate existence of the ma-  terial world. All this is Knowledge, ac-  quired indeed so early in life, that its com-  mencing and progressing steps are forgotten ;  but we are nevertheless warranted in affirm-  ing that not the least part of it, is an original  gift of nature. Along with this knowledge  we acquire emotions and passions ; for to knoia  material objects, is to know them as causes of  pleasurable or painful sensation, and hence to  feel for them, in various degrees, and with  various modifications, desire and aversion, joy  and grief, hope and fear. And here, as the  same object does not always produce the same  emotion, or the same emotion arise from the  same object, we begin a new class of abstrac-  tions : we separate, mentally, the object from  the emotion or the emotion from the object :  we are enabled in consequence to abstract and  consider those differences in the objects, from  which the different effects arise, and to ascer-  tain, by trial, how far they yield to volition ope-  rating by the exterior bodily members, which     SECT. 6.3     ON GRAMMAR.     19     we have previously discovered to be subservient  to the will. In this new class of abstractions,  and the consequences which arise from them,  we shall find the beginning of that knowledge  which human reason is privileged to obtain,  compared with that which the higher orders  of the brute creation in common with man,  are able to reach j and from this point we  shall be able to trace how man becomes /ie'poyjr,  or divider of a natural word into parts of  speech *, while other animals retain unaltered  the cries by which their desires and passions  are first expressed.   6. As we are able to separate, mentally,  the object from the emotion, and to remem-  ber the natural cry after the occasion that  produced it ceases, the natural cry might re-  main as a sign either of the object or of the  emotiont. But this does not carry us beyond   • Thia is the sense in which we choose to under-  stand the word, and not merely voice-dividing or ar-  ticulating.   ■f For instance, as, in the present state of language,  the exclamation of surprise ha-ha '. is either an inter-     to     ON GRAMMAR.     the mind which forms the abstraction, and  has the power to establish a sign (wliether  audible or not) to fix and remember it: — our  inquiry is, how a communication can be made  from mind to mind, when the signs which na-  ture furnishes are inadequate to the occasion.  And first be it observed, that only such occa-  sions must, at the outset, be imagined as do  but just rise above those for which the cries  of nature are sufficient: — we must not sup-  pose a necessity for communicating those ab-  stract truths which grow out of an improved  use of language, and which could not there-  fore yet have existence in the mind. And  we have further to observe that no com-  munication can be made from one mind to  another, but by means of knowledge which  the other mind possesses; — the cries of na-  ture can find their way only into a conscious  breast, — that is to say, a breast that has known,     jection eignifyiDg that emotiou, or the n  so placed ae to give occasion to it.     SECT. 6.3 ON GRAMMAR. 91   or at least can know, the feelings which are  to be communicated, and is capable, therefore,  of sympathy or antipathy ; and knowledge  of whatever kind can be conveyed to another  mind only by appealing to knowledge which is  already there. To suppose otherwise, would  be to attribute to human minds what has been  imagined of pure spirits, — the power of so  mingling essences that the two have at once  a common intelligence. To human minds It  is certain that this way of communicating is  not given, but each mind can gain knowledge  only by comparing and judging for itself, and  to communicate it, is only to suggest the sub-  jects for comparison. Let us suppose that a  communication is to be made for which a na-  tural cry is not sufficient, — the difficulty, then,  can be met only by appealing to the know-  ledge which the mind to be informed already  possesses. The occasion will create some cry  or tone of emotion ; but this we presuppose  to be insufficient. It will however be under-  stood as far as the hearer's knowledge may     02 ON GRAMMAR, [CIIAP. 1.   enable him to interpret it — that is, he will  know it to be the sign of an emotion which  himself has felt, and he will think perhaps of  some occasion on which himself used it. But  the cry is to be taken from any former par-  ticular occasion, and applied to another; and  he who has the communication to make, will  try to give it this new application by joining  another sign, such as he thinks the hearer is  hkewise acquainted with. The natural cry  thus taking to its assistance the other sign, and  each limiting the other to the purpose in hand,  they will, in their united capacity, be an ex-  pression for the exigence, and will, to all in-  tents and purposes, be a sentence.   7. In some cases, nature seems to furnish  an instinctive pattern for the process here de-  scribed : —a man cries out or groans with pain ;  he puts his hand to the part affected, and we  at once interpret his cry more particularly  than we could have done without the latter  sign. In other cases, we are driven to the  same process not by an instinct, but by the      SECT, 7*3     ON GEAMMAH.     ingenuity of reason seeking to provide that  which nature has not furnished. If a man  unskilled in language, or not using that which  his hearers understand, should try to make  known what art expresses by a sentence such  as " I am in fear from a serpent hidden there,"  his first effort would be the instinctive cry of  fear ; but aware that this could be particularly  interpreted only of a known, and not of an un-  known occasion, he would, by an easy effiirt of  ingenuity, fix it for the present purpose by add-  ing a sign or name of the reptile, (for mimick-  ing the hiss of the reptile would obviously be  a name,) and by joining to both these a ges-  ticulative indication of place. The instinctive  cry thus newly determined and appUed, is a  sentence ; and however clumsy it may seem  when compared with the more complicated  one previously given, yet the art employed is  of the same kind in both. We leave the read-  er to smile at the example as he pleases, and  will join in his smile while he compares it with  that in the epistle of the poet in the note at     ^     ON GRAMMAR.     I^CHAP. I.     Sect. 1.; and, if he is disposed to smile again,  we will suppose another example : — Two men  going in the same direction, are stopped by  an unexpected ditch, and ejaculate the na-  tural cry of surprise ha-ha/ This is remem-  bered as the expression suited for that par-  ticular occasion; and the mind, the human  mind, seems to have the power of generalizing  it for every similar object. Suppose one of  these men finding another ditch very offensive  to his nose, signifies this sensation by screwing  up the part offended, an d uttering the nasal  interjection proper for the case ; — the inter-  jection may not be sufficient j for the other  man may remain to  be informed of what his  companion knows, namely that the offence  proceeds from the ditch. To fix the mean-  ing, therefore, o f the interjection to the case  in hand, the communicator adds the former  natural cry in order to signify the ditch, and  the two signs qualifying each other, are a  sentence.   8. An artificial instrument as language is,     SECT. S.J     ON GRAMMAR.     25     growing (as we suppoaej out of necessity, and  adapted at first to the rudest occasions ; per-  fected by degrees, and becoming more com-  plicated in proportion as the occasions grow  numerous and refined ; — such an instrument,  when we compare its earliest conceivable state  with that in which it  has received its iiighest  improvement, must appear clumsy and awk-  ward in the extreme. But in the very rude  state in which we here suppose it, the art em-  ployed is essentially the same as afterwards :  — two or more signs are joined together, each "  sign referring separately to presupposed know-  ledge, but in their united capacity communi- i  eating what is supposed to be unknown. Of  the signs used, that must be considered the ,  principal by which the speaker intimates the ,  actual emotion j the other signs, which do but j  fix its meaning, are secondary. Thereforej ;  though the appellation word (that is p^/io, i  dictum, or communication,) strictly belongs  to the whole expression or sentence, we may  reasonably give that appellation to the prin-     Sfi ON GRAMMAR. [CHAP. I.   cipal sign. According to this supposition,  the original verb was an expression equiva-  lent to what we now signify by I hunger, I  thirst, I am warm, I am cold, I see, I hear,  IJeel, &c., / am in pain, I am delighted, I am  angry, 1 love, I hate, I fear, I assent, I dis-  sent, I command, I obey, &c. Whether this  a priori conjecture has any facts in its favour,  is an inquiry suitable to the etymologist, but  fo reign to our purpose, because, whether true  or not, the general argument by which we in-  tend to prove the nature of the parts of speech,  will remain the same*.     " Vet it may be worth while to quote the coinci-  dent opinion of another writer. " It may be asked "  says Lord Monboddo, " what words were (irst invented.  My answer is, that if by words are meant what are  commonly called parts of speech, no words at all were  first invented ; but the first articulate sounds that were  formed denoted whole sentences ; and those sentences  expressed some appetite, desire, or inclination, relating  either to the individual, or to the common business  which I suppose must have been carrying on by a herd  of savages before language was invented. And in this     SECT. 9.3     GHAMMAR.     S7     9. We have next to imagine the use of  any of the foregoing verbs in the third per-  son ; for that, it should seem, would be the  next step. In communicating that anothet-  hungers or thirsts, or sees or hears, or is angry  or pleased, &c., the difficulty would be to give  the word this new application, and a limiting  sign would, as usual, be necessary. A proper  name would be the sign required ; and if not  too great a tax upon fancy, we may conceive  the invention of these from the mimicking of a  man's characteristic tone, or his most frequent  cry ; not to mention the assistance of gesticu-  lative indication. But when verbs had thus  lost the reference which, at first we presume,  they always bore to the speaker, a sign,  whether a change of form, or a separate word,  would be wanted to bring them back to their  early meaning as often as occas ion required.  A gesticulative indication of the speaker and     way I believe language continued, perhaps for many  ages, before names were invented." — Origin and Pro-  grese of Language. Vol. I. Book 3. Chap. 1 1-     28 ON GRAMMAR. [CHAP. I.   of the person spoken to, can easily be con-  ceived : how soon tliese would give place to  equivalent audible signs, the reader is left to  calculate j and as to the pronoun of the third  person, he may allow a longer time for its in-  vention, especially as even in the finest of lan-  guages, tliere is no word exactly answering to  ille in Latin and he in English.   10. We have suggested a clew to the in-  -yention of proper names, and (for the reader  jnust allow us much) we will suppose these,  L ^ far as need requires, to be invented. But  r piost of these, from the difficulty of inventing  a new name for every individual, would gra-  dually become common. If a man has called  I the animal he rides on by a proper appellation  I corresponding to horse, what shall he call  t Other animals that he knows are not the same;   and yet resemble? Because he is unprovided ..  r jwith a name for each individual, he will call'  I each of them horse*, and the name will then   " Compare Adam Smith, " Considerations con-  cerning the First Formation of Languages," appended     SECT. 10.] ON GRAMMAR.   no longer be proper but common. But the  same powers of observation which acquaint  us with the points of resemblance, likewise  show the points of difference, and when we  wish to distinguish the animals from each  other, how is this to be done ? The question  is easily answered when we have a perfect lan-  guage to refer to, but it was a real difficulty  when the expedient was first to he sought.  Yet the difficulty not unfrequently occurs  even in a mature state of language, and the  manner in which it is overcome, will enable  us to conceive how, in the rude state of Ian-  guage we are supposing, itwas universally met,  till the noun-adjective became a part of  speech*. Of two horses, we observe that one   to his work on the Theory of Moral Sentiments. As a  proof how prone we are to extend the appellation of  an individual to others, he remarks that " A child just  learning to speak, calls every person who comes to the  house its papa or its mamma ; and thus bestows upon  the whole species those names which it had been taught  to apply to two individuals."   ' The Mohegans " (an American tribe) " have     so ■ ON GRAMMAR. [cHAP. I.   has the colour of a chestnut, and the other is  variegated hke a pie ; and we call the former  a cfieslnut horse, and the other a pied or piebald  horse. Here we perceive are two nouns-sub-  stantive joined together to signify an indivi-  dual object, and employed, Ui their united ca-  pacity, to signify what would otherwise have  been denoted by an individual or proper name.  This, then, is their meaning, respectively,  as a single expression. In their abstract or  separate capacity, the one word denotes either  one or the other of the two animals without  reference to the difference between them : the  other word denotes, not a chestnut or a pi^  but that colour in a chestnut, and those varie-  gated colours in a pie, by which one of the  animals is distinguished from the other, and  these words are no longer nouns-substantive  DO adjectives in all their language. Although it may  at first seem not only singular and ciuious, but im-  possible that a language should exist without adjectives,  yet it is an indubitable fact," — Dr. Jonathan Edwards  — quoted by H. Tooke, Diversions of Purley, Vol. II.  p. 463.     SECT. 10.]     ON GBAMMAR.     but nouns-adjective *. And here the ques-  tion will naturally occur, how would a hearer  know when a noun was used substantively,  and when adjectively ? As this would often  be attended with doubt and ambiguity, the  necessity of the case would soon suggest  some slight alteration in the word as ofi;en as  it was used adjectively ; and the same all-  powerful cause would likewise, in time, dia-  tinguish adverbs from adjectives : for at first  an adjective would be used without scruple to  limit the verb, as to limit the substantive j since     • " The invention of the simplest nouns-adjective,*'  says Adam Smith, " must have required more meta-  physics than we are apt to be aware of." But the dif-  ficulty he imagines is done away by the hypothesis  suggested above ; and how near it is to the truth, will  fae conceived by calling to mind the ready use of al-  most any substantive as an adjective, as often as need  requires : e. g. a chestnut horse, a horse chestnut ; a  grammar school, a school grammar ; a man child, a  cock sparrow, an earth worm, an air hole, a (ireking,  a water lily ; not to mention the innumerable com-  pounds that are considered single words ; as, seaman^  Iiorsenian, footman, inkstand, coalhole, bookcase, Sic.     «t     ON GRAMMAR.     [chap. 1     this is often done even in the present state of  language j but the doubt whether it was to be  taken with the substantive or the verb* would  soon produce some general difference of form ;  and thus the adverb would be brought into  being as a distinct part of speech.   11. Still it would often happen, that in  endeavouring to limit a verb to the particular  communication in view, no substantive or pro-  noun joined to it, not even with the further  aid of an adjective or adverb joined to the  substantive or verb, would suffice ; and failing,  therefore, to convey the communication by  one sentence, it would become necessary to  add another to limit or determine the significa-  tion of the first. Now a qualifying sentence  thus joined, when completely understood in  connexion with that it was meant to qualify,  would be esteemed as a part of the same sen-  tence, and the verb, in the added sentence,     • E. g. whether " I love much society " is to be  understood / much-li/ve suciety, or, / Iwe 7iutch-  society.     SECT. 11.] ON GRAMMAR. 38   would possibly then lose its force as the sign   of a distinct communication. This again, will  easily be understood by a reference to what  occurs in the present state of language. Look-  ing at the sentence, " In making up your par--  ty, except me," no one hesitates to call concept  a verb ; but in this sentence, *^ All were there,  except me," although the word except has pre^^  cisely the same meaning, yet, as we do not con^  sider the clause except TTie to be a distinct com-  munication, but only a qualification to suit the  whole sentence to the purpose in view, we call  except a preposition *, that is, a word put be^     * This solution of the difficulty in the invention  of prepositions, which seems so considerable to Adam  Smith, is suggested, as the reader will perceive, by  the etymological discoveries of Home Tooke, and will  receive complete confirmation by the study of his ad-  mirable work. Let it not be supposed, however, that  we have nothing to object to in the Diversions of  Purley : some ftmdamental principles we have already  marked for inquiry ; and on the point before us, we  have to observe on that curious way of thinking, which  leads him, because a word was once a verb or a noun.     Olf GRAMMAR.     j^CHAP. I.     fore another to join it to the sentence that  goes before.   12. But in thus qualifying sentence by sen-  tence, it may sometimes be necessary to use  three verbs, one of them being merely the sin-  gle verb that joins the two sentences together ;  as, " I was at the party, and (i. e. add, or join  this further communication) I was much de-  lighted." Sometimes a noun will be used in  this way ; as, " I esteemed him, because (i. e.  this the cause) I knew his worth." Any par-  ticular form of verb or noun used frequently  in this manner to join sentence to sentence,  will cease at last to be considered any thing  more than a conjunction *.   IS. As to the article, we have only to sup-  to esteem it always so ; on the same principle, no doubt,  that, because the word truth comes from he trou-eth or  thinkelh, a.aA a man's thoughts are always changing,  he denies that there is any such thing as eternal, im-  mutable truth.   * Again the reader is referred to the Diversions of  Purley, for a confirniation of this account of the birth  of conjuncticms.     SECT. 14,] ON GRAMMAR. 85   pose some adjective used in a particular limit-  ing sense so frequently, that we at last regard  it as nothing more than a common prefix to  substantives : — as to a participle^ it is confess-  edly, when in actual use, either a part of the  verb, or a substantive, or an adjective : — and  as to an interjection^ this we have supposed to  be the parent word of the whole progeny ; and  if it is sometimes used among the parts of  an artificial sentence, it is only as a vibration  of the general tone of feeling that belongs to  the whole.   14. In this manner, or in a manner like  this in principle and procedure, would lan-  guage grow out of those powers bestowed on  man by his Creator, even though it had not  been directly communicated from heaven :-—  in this manner is the progress from natural  cries to artificial signs contemplated and pro-  vided for by the constitution of the human  mind; — in this manner would the parts of  speech be developed j and men placed in so-  ciety, and endowed with powers for observa-   D S     36 ON GRAMMAR. [CHAP. I.   tion, reflexion, comparison, judgment, would,  in time, become fiepoire^f or dividers of a na-  tural word into significant parts, with the  same kind of certainty that they become bipeds  or walkers on two legs* ; being bom neither  one nor the other.   * And according to Monboddo, with the same  certainty that they lose their tails; for when they  were mutu/m, et turpe pecus^ he appears to think  they might have been so appendaged ; nay, he knew a  Scotchman that had a tail, though he always took care  to hide it : (his lordship was surely in luck^s way to  find it out.) After all, it would be difficult to prove,  notwithstanding the authorities Monboddo quotes, that  herds of men were ever found destitute of language.  Leaving, therefore, the origin of the first language,  and the subsequent confiision or division of it precisely  as those two &ct8 stand in Genesis, all we mean to  assert in the text is this, — that if a number of children  having their natural faculties perfect, were suffered to  grow up together without hearing a language spoken,  they would invent a language for themselves : though,  for a long time, it might remain nothing better than  that of the Hurons described by Monboddo, (Origin  and Progress of Lang. VoL I. Book 3. Chap. 9.) in  which the parts of speech are scarcely evolved, from  the original elements, but what in a formed language     SECT. 15.] ON GRAMMAR. S^   15. But the object of the foregoing at-  tempt, was not so much to trace the origin     is expressed by several words, is expressed by a sign  not divisible into significant parts. Thus, he says,  there is no word which signifies simply to cut, but many  that denote cuttingjish^ cutting wood^ cutting chaths,  cutting the heady the arm^ &c. And so of the language  throughout. More than one generation would be re-  quired, and very favourable stimulating circumstances,  to bring such a chaos of a language into form ; but  that the human mind has within itself the powers for  accomplishing it sooner or later, we see no cause to  doubt — These words, and the whole of the hypothesis  in the text above, were written before the third Volume  of Dugald Stewart's Philosophy of the Human Mind  had been seen. From that part which treats on Lan-  guage we quote the following passages :   ^^ That the human faculties are competent to the  formation of language, I hold to be certain.*"   ^^ Language in its rudest state would consist partly  of natural, partly of artificial signs ; substantives being  denoted by the latter, verbs by the former.*"   These are among the many passages which coincide  with the views opened in the previous hypothesis. It  is to be added, that D. Stewart considers the imperative  mood to be the first form in which the artificial verb  would be displayed.     ON GRAMMAR.     [chap. 1.     and first progress of language, as to get at  the real ground of diflference among the se-  veral parts of speech. On this subject, there  prevails a universal misconception. Prom the  definitions and general reasoning in Gram-  mar ; — from the theories laid down in Logic ;  — and the basis on which the rules and prac-  tice of Rhetoric are presumed to stand, this  principle seems to be taken for granted, that  the parts of speech have their origin in the mind  independently of the outward signs, when, in  truth, they are uothing more than parts in the  structure of language ; contrivances adopted  at first on the spur of theoccasion, the shifts  and expedients to which a person is driven,  ■when not being able to lay bare his mind at  once according to his consciousness, he tries,  by putting such signs together as were used  for former occasions and therefore known as  regards them, to form an expression, which, as  a whole, will he a new one, and meet the pur-  pose in hand. True indeed it is, that these  very contrivances become, in their more re-     SECT. 15.2 ON GRAMMAR. 39   fined use, the great instruments of hmnan rea-  son by which all improvement, all extensive  knowledge, is obtained; but we are not to  confound the instrument with the intelli-  gence that uses it/ nor to suppose that the  parts of which it is composed, have, of ne-  cessity, any parts corresponding with them in  the thought itself. It is not what a word signi-  fies that determines it to be this or that part  of speech, but how it assists other words in ma-  king up the sentence. If it is commissioned to  unite the whole by the reference immediate  or mediate which all the other words are to  bear to it, and to signify that they are a sen-  tence, that is, the sign of a purposed commu-  nication, then it is the verb : — if it has not  this power, (namely, of uniting the other words  into a sentence,) and yet is capable, in all other  respects, of standing as an independent sign,  (this sign not being the sign of a purposed  communication) then it is a substantive .-—if it  is the implied adjunct of a substantive, it is an  adjective or an article^ — if of a verb^ an ad-     40 ON GRAMMAR. [CHAP. I.   verb : — if we know it to be a word, which, in  a sentence, is fitted to precede a substantive,  (or words taken substantively) in order to con-  nect such substantive with -what goes before,  then it is a preposition : — and if it goes before,  or mingles in a sentence, in order to connect  it with another sentence, then it is a conjunc-  tion. These are the only real differences of  the parts of speech : — as to the meaning, that  does not of necessity differ because a word is  a different part of speech ; — the following  words, for instance, all express the same no-  tion :   Add   Addition   Additional   Additionally   With*   Andt   * The imperative of the Saxon verb Jpi^an to join.   -|- The imperative of the Saxon verb ananab to add.   The place and ofHce of these six words in a sentence  would of course differ, and the sentences in which they  were respectively used would require a various arrange-     SECT. 15.] ON GRAMMAR. 41   Our definitions reach the real differences  among these words, and they will be found  adequate to all differences, when, by the ob^  servation hereafter to be made, we are quali-  fied to make due allowance for the licences  assumed by the practical grammarian *• In   ment to meet the same purpose, but as to the meaning  of the words, it would be the same in whatever  sentence : e. g.   Add something to our bounty.   Make an addition to our bounty.   Give an additional something to our bounty.   Give additionally to our bounty.   Increase o ur bounty with the gift of something.   Consider our bounty and give likewise.  * To suit our definitions to an elementary grammar,  they must be quaUfied and circumstanced: — a verb,  for instance, must be shewn to be a word that is by  itself a sentence, as esurio ; or which signifies a  sentence, as I am hungry ; or which is fitted to sig-  nify a sentence, as am, lovest. A verb in the infinitive  mood, is a verb named but not used ; a8 to be, to love ;  or if used in a sentence, it is not the verb. A noun-  substantive is a name capable of standing independently,  but it cannot enter into a sentence except by being  connected directly or indirectly with a verb. The in-  flexion of a noun-substantive, as Mard, Mark'' 8^ is     ON GRAMMAS.     [chap. I,     the mean time, in order to throw as much  light as possible on the nature of the con-  nexion between thought and language, let us  look back a little on foregoing statements,  and partially anticipate those which are to be  opened more at full under the heads of Logic  and Rhetoric.   called a substantive, bnt in so calling it, we must say  a Bubstantive in the genitive, or other case. A noun-  adjective is a name not fitted to stand independently,  but to be joined to a noun-substantive, and so to form  with it one compound name. An adverb is a word not  fitted to stand independently, but to be joined to a verb,  and to form with it one compound verb, A preposition  ig a word governing as its object a substantive or pro-  noun in the manner of a verb, but not an obvious part  of a verb, nor capable, like a verb, of signifying a  sentence. The article, pronoun, participle, conjunc-  tion, and interjection, may be defined as usual. We  would suggest moreoverthat in an elementary grammar,  no definition, and no part of a definition, should  be brought forward, till absolutely required by the  examples that are immediately to follow it. In  teaching a child, it is the greatest absurdity in the  world to set out with general principles, when the  business is, to reach those principles by the eiiamina-  tion of particulars.     SECT. 1(3-3 '^^ GRAMMAB. 43   1 6. It may be that the organs of sensation  are not all fully developed in a new-born in-  fant ; but if, for the sake of our argument, we  allow that they are so, this is as much as to  say, that our earliest sensations from the ob-  jects of the material world, are the same that  they are afterwards. But there must be this  most important difference, — that the early  sensations are -wilkoui knowledge, and the lat-  ter, with it. I know that the object which now  affects my sense of vision, is a being like my-  self, — I know him to be one of a great many  similar beings ; — I know him to be older or  younger than many of them, — to be taller or  shorter; — I know pretty nearly the distance  he is from me ; — 1 know that the particular  circumstances under which he is now seen,  are not essential to him, but that he may be  seen under other circumstances : — I know that  what now affects my sense of hearing, is the  cry or bark of a dog j — I know, although my  eyes are shut, that there are roses near me,  or something obtained from roses j — I knoie     u     ON GRAMMAR.     [CUAP. ]     that sometliing hard has been put into my  mouth ; — and now I know it to be part of an  apple. All the sensations by which the  various knowledge here spoken of is brought  before the mind, the new-born infant may  possibly be capable of; but as to the know-  ledge, there is no reason to believe he lias the  least portion of it. For the knowledge is  gained by experience, requiring and com-  prising many individual acts of observation,  comparison, and judgment j all which we  suppose yet to take place in the new-born  infant. Now, in looking back to what has  been said on the acquirement of language, we  find the effect of our progressing knowledge  to be this, that every sign arising out of a par-  ticular occasion, will lose that particular re-  ference in proportion as we find it can be used  on other occasions j and so all words will, at  last, in their individual capacity, become ab-  stract or general. This is as true of such  words as yellow, white, heat, cold, soft, hard, .  bitter, sweet, and the like signs of what Locke     SECT. 16.] ON GRAMMAR. 45   calls simple ideas as of any other * : for we  can evidently use these words on an infinity  of different occasions j and the power of so  using them is an effect and a proof of our  knowing that the different occasions on which  we use the same word, have a something in  common, or in some way resemble. But  while all words thus acquire an abstract or  general meanipg, every communication which  we purpose to make by their means, must, in  comparison with their separate signification,  be particular ; and our putting them together  in order to form a sign for the more particular  thought, will be to deprive them of the abstract  or general meaning which they had indi-  vidually. If this is the real nature of the  process, we are completely mistaken if we  suppose that every word in a sentence sig-  nifies a part of the whole thought, and that  the progression of the words is in corre-  spondence with a progression of ideas which  the mind first puts togetlier within, and then  * Vide Locke, Book II. Chap. 1. Sect. 3.     46 ON GRAMMAR. j^CHAP. I.   signifies without What deceives us into this  impression, is, that on considering each word  separately, each is found to have .1 meaning.  Let us try, however, whether the joining of  words into a sentence, does not take from them  the meaning they have separately. Put to-  gether the three words " My head aches,"  and we have an expression, namely the whole  sentence, which signifies what, from a want  of clearness in our remarks, may possibly be  the reader's present particular sensation: hut  my, separately, signifies the general knowledge  I have attained of what belongs to ine as dis-  tinguished from what belongs to another j a  knowledge which is not at all necessary (that  is, the ^'•CTJcra/ knowledge) to the sensation it-  self, nor even to the expression ofit, if we could  find any single sign in lieu of the three which  we have put together. Accordingly, the  word my, as soon as it is joined to the other  words, drops that meaning which it had  separately, and receives a particular limitation  from the word head, which word head is like-     SECT. 17.3 OM GRAMMAR.     47     wise limited by the word rrof ; and the more  particular meaning which both these receive  by each other, is limited to the particular oc-  casion by the word aches. Yet, it may perhaps  be thought, that in this, and in every other  sentence, each word, as the mind suggests it  to the lips, is accompanied by the knowledge  of its separate meaning, and that, in this  manner, if we use the word idea in the un-  restricted sense familiar to the readers of  Locke, each word may be said to represent an  idea. Without entirely denying the justice  of this view of the matter, we offer in its place  the following statement :   17. In forming a sentence for its proper  occasion, the knowledge of which each sepa-  rate word is fitted to be the sign, may, or  may not be in the mind of the speaker: it  may be entirely there, or only in part, or not  at all there ; that is to say, the speaker may  not know the separate meaning of a word,  but only the meaning it is to have in union  with the other words. And even if the     48 ON GRAMMAR. [CHAP. I.   speaker does know the full separate meaning  of each word, yet he is not under the neces-  sity of thinking of that separate meaning  every time he uses it : nor does he, in fact,  think of the separate meaning of words while,  in putting them together, his purpose is to ex.  press what has been often expressed before, but  only (and even then but partially and occa*  tonally) when he uses words to work out some  conclusion not yet established in his own mind,  or when a train of argument is required to  convince or persuade other minds. This  statement will of course require some con-  siderations in proof.   18. And first, as to the knowledge of  which each separate word is fitted to the sign,  it is to be observed that our knowledge grows  with the use of words, and therefore our firet  use of them is unaccompanied by that know-  ledge which we gain by subsequent use.  This is true, whether we invent words, or  adopt those already invented. In the rude  beginning of language, the first use of a word     SECT. IS.J ON GRAMMAH.     4.9     for head, would be a use of it for a particular  occasion, and the word would be particular or  proper. If the speaker used it with reference  to himself, it would signify what we now sig-  nify fay the two words my head ". By observ-  ation and comparison, he would find he could  extend the meaning of the word, and apply it  with reference to his neighbours as well as  himself, and it would then no longer be proper  but common ; that is to say, it would signify  a human head, and not mj/ head. Extending  his observations still more widely, he would ap-  ply it with reference to every other living crea-  ture, and it would accordingly then signify a /(u-  ing creature's head. Looking and comparing  still further, he would apply it with referenceto  every object, in which he discovered a part  having the same relation to the whole as the  head of a living creature has to its remaining  parts ; and the word would then, and not till  then, have its present meaning ; that is to   " Compare the characteristics of the Huron lan-  guage referred to in the note appended to Sect. 14.     ON GRAMMAR.     [chap.     say, in a separate unlimited state it would  signify neither my head, nor a human head,  nor a living creature's head, but the top,  chief part, beginning, supremacy of any  thing whatever. Nor is the process essentially  different in acquiring the use of words already  invented. A child does not at first put words  together, but, if his head aches, he will say  perhaps "head! head!" using the single  word in place of a sentence. At length he  will say mi/ head, and brother's liead, and  horse's head, and cradle's head. Still there  are other applications of the word to be  learned by use ; and it surely will not be  contended that any one knows the meaning  of a word beyond the cases to which he can  apply it. The knowledge which a separate  word is fitted to signify, may then be wholly  or may be partly in the mind of him who  uses it in a sentence ; and it is very possible  not to be there at all. A foreigner, for in-  stance, who had beard the phrase the head of  the army applied to the general-in-chief,     SECT. 19.3 ON GRAMMAR. 51   would know the meaning of the phrase, but  might be quite ignorant of the meaning of  the separate words, or even that it was com-  posed of separable words : and probably most  people can look back to a time in early life,  when they were in the habit of using many a  phrase with a just application as a whole,  without being aware that it was reducible  into parts in any other way than as a poly-  syllabic word is reducible.   ig. But even when the speaker, in form-  ing a sentence, has previous possession of all  the knowledge of which each word is sepa-  rately fitted to be the sign, yet he does not in  general think of their separate meaning while  he is putting them together, but only of the  meaning he intends to express by the whole  sentence. For through the frequent use of  phrases and sentences whose forms are hence  become familiar, there is scarcely any senti-  ment, feehng, or thought, that suddenly arises  in the mind, that does not as suddenly sug-  gest an appropriate form of expression. This     [chap.   is manifestly the case with such sentences as  arc in constant use for common occasions :  these the speaker cannot be said to make,  they occur ready-made, and he pronounces  the words that compose them with as little  thought of their separate meaning as if he  had never known them separate. Even when  sentences ready-made do not occur, yet the  forms of sentences will occur, and the speaker  will, in general, do nothing more than insert  new words here and there till the sentence  suits his purpose. Thus he who had said  " My head aches," will recollect the form of  sentence when his shoulder aches, and in  using the sentence, will only displace head  for shoulder: or if his head " is giddy," he  will only displace aches for the two words  quoted, in order to say what he feels.   20. When indeed we use language for  higher occasions than the most ordinary in-  tercourse of life ; when by its means we pro-  secute our inquiries after truth, or use it dis-  cursively as an instrument of persuasion, then     5ECT. 20.] ON GRAMMAR. 63   the operation itself is carried on by dwell-  ing on and enforcing the abstract mean-  ing of some of the words and some of the  phrases whUe in their progress towards form-  ing sentences, as of the sentences while in  their progress toward forming the whole ora-  tion or book. But in such cases, language  may more properly be said to help others to  come at our thoughts , than to represent our  thoughts : although it is likewise true, that  we could not ourselves have come at them  but by similar means. Independently of the  words, therefore, the thoughts would have  had no existence j neither should we have  proposed the inquiry after the truths we seek,  nor have imagined any thing in other minds,  by addressing which they could be influenced.  Still, however, in these higher uses of lan-  guage, (uses which are to be dwelt on more  at full in the chapters on Logic and Rhe-  toric,) there is the same difference between  words separately, and the meaning they re-  ceive by mutual qualification and restriction ;     «*     ON GRAMMAR.     [chap. I.     that is to say, in these higher uses of lan-  guage, 83 well as in those already remarked  upon, the parts that make up the whole ex-  pression, are parts of the expression in the  same manner as syllables are parts of a word,  but are 7tol parts of the one whole meaning in  any other way than as the instrumental means  for reaching and for communicating that  meaning. And suppose the communication  cannot be made but by more signs than use  will allow to a sentence, — suppose many sen-  tences are required — many sections, chapters,  books, — we affirm that, as the communica-  tion is not made till all the words, sentences,  sections, &c. are enounced, no part is to be  considered as having its meaning separately,  but each word is to its sentence what each  syllable is to its word ; each sentence to its  section, what each word is to its sentence ;  each section to its chapter what each sen-  tence is to its section, &c. Thus does our  theory apply to all the larger portions of dis-  course, and to the discourse itself, Aristotle's     SECT, 20%]] ON GRAMMAR* 55   definition of a word, namely, ** a sound sig.  niiicant. of which no part is by itself signi^  ficant ;" * for if our theory- is true, the words  of a sentence, understood in their separate  ^rapacity, do not constitute the meaning of the  whole sentence, (i. e. are not parts of its  whole meaning,) and therefore, as parts of  that sentence, they are not by themselves  significant ; neither do the sentences of the  discourse, understood abstractedly, constitute  the meaning of the whole discourse, and  therefore, as parts of that discourse, they are  not by themselves significant : they are sig-  nificant only as the instrumental means for  getting at the meaning of the whole sentepce  or the whole discourse. Till that sentence  m oration is completed, the Word t is unsaid  which represents the speaker's thought- If   ♦ 4^6jvii (ni/xAVrixiii vi'; A*sf oj oOih B<rri xalP abrh arif/iotv-i   rikiv. De Poetic c. 20.   f In this wide sense of the expression is the Bible  called the Word of God. We shall distinguish the  term by capitals, as often as we have occasion to use it  with simitat comprehensiveness erf meaning.        '$^ ON GRAMMAR. [ CHAP. I.   it be asserted that the parallel does not hold  good with regard to such words as Aristotle  has in view, because, of words ordinarily so  called, the parts, namely the syllables, are not  significant at all, while words and sentences  which are parts of larger portions of dis-  course, are admitted to be abstractedly sig-  nificant, however it may be that their abstract  meaning is distinct from the meaning they re-  ceive by mutual limitation, — we deny the  fact which is thus advanced to disprove the  parallel : we affirm that syllables are signifi-  cant which are common to many words ; for  instance, common prefixes, as wn, mis, corif  dis, bi, tri, &c.; and common terminations,  as nesSjJul, hood, tion, fy, &c. j and so would  every syllable be separately significant, if it  occurred frequently in different combinations,  and we could abstract out of such combina-  tions the least shade of something common in  their application : nor is it peculiar to syllables  to be without signification individually; the  same thing happens to words when they are     \     SECT. 21.] ON GRAMMAR. 57   always combined in one and the same way in  sentences *. Conceiving, then, that we are  fully warranted in the foregoing statement, we  affirm it to be the true basis of Grammar, Lo-  gic, and Rhetoric. Leaving the latter two  subjects for their respective chapters, we pro-  ceed, in this chapter, with such further proofs  as may be necessary to confirm our position  as far as Grammar is concerned.   21. We have imagined the gradual de-  velopment of all the parts of speech recog-  nized by grammarians ; but no reference has  yet been made to the inflexions which some  of them undergo; nor to the diflference of  meaning they receive in consequence of such  inflexion ; nor to interchanges of duty among  the several parts of speech ; nor to pecu-  liarities of use, which so oflen take from them  their characteristic differences; nor to va-   " What separate meaning, for instance, is there,  now, in the words which compose such phrases as, by-  and'bij, goodJi'ye, ftatc-du-you-do, 8cc.     I ON GEAMMAB. t^CHAP. I.   riety of phrase in expressing the same mean-  ing j nor to the power which we frequently  exercise of making the same communication  by one or by several sentences ; nor, in  short, to the multitude of refinements which  grow out of an improving use of language,  many of which seem to confound and destroy  the definitions we obtain from the first and  simplest forms of speech. All these seeming  irregularities will, however, find a ready key  in the general principles we have ascertained.  For our general principles are these : i. That  two or more words joined together in order  to receive, by means of each other, a more  particular meaning, are, with respect to that  meaning, inseparable j since, if separated,  they severally express a general meaning not  included in the more particular one. Hence  it follows, that words may as easdy receive a  more particular meaning by some change of  form, as by having other words added to  them : nay, it seems more natural, when the  principle is considered, to give them a more     SECT. 21.]     ON GRAMMAR.     59     particular meaninjj by a change of form than  fay any other way. — ii. That a word is tliis or  that part of speech only from the. office it  fulfils in making up a sentence. From this  principle it follows, that a word is liable to  lose its characteristic difference as often as it  changes the nature of its relation to other  words in a sentence ; and it also follows, that  every now and then a word may be used ia   L8ome capacity wliich makes it difficult to be  assigned to any of the received classes of  words. — iii. That since the parts of which a  sentence is composed denote general know-  ledge, distinct from the more particular mean-  ing of the whole sentence, it may be possible i  to work our way to a particular conclusion,  either in reasoning for ourselves or in per- j  auading others, by putting such words to-  gether as form a sentence, that, as a whole,  expresses the particular conclusion; but that  when, from the length of the process, this  cannot be accomplished in a single sentence,  we shall be obliged to work our way by many     ON GRAMMAR.     [chap. I,     sentences, whicli will bear the same relation  to the conclusion implied by them as a whole,  as the parts of each sentence bear to what  the sentence expresses. From this principle  it follows, that using many or fewer sentences  to arrive at the same result, will frequently  be optional. The examination of these se-  veral consequences a Httle more in detail  with reference to the principles from which, i  they flow, will complete the chapter.   22. It is well known, that the inflexions  which nouns, verba, and kindred words are  liable to in many languages, are comparatively  unknown in English, the end being for the  most part attained by additions in the shape  of distinct words. Thusthe particular re-  lation of the word Marcus to the other words  in the sentence, which in Latin is made  known by altering the word into Marco, is  signified in English by the word io ; and to  MarcuSy esteeming the two words as one ex-  pression, is the same as Marco. So likewise  the word amo, which in English signifies /     SECT. 23.] ON GEAMMAR.     Gl     l&ve, is adapted to a different meaning by  being changed into amabit, which in English  is to be signified by he mil love, the three  words, taken as a whole, being the same as  the single Latin word. Shall we call to Mar-  cus the dative case of Afarcus, and he will ,  love, the third person singular of the future  tense of / love, as Marco and amabit are re-  spectively called with reference to Marcus  and amo? or shall we parse (resolve into  grammatical parts) those English sentences,  and so deny, in our language, a dative case and '  a future tense ? It is evident that this is a  question which only the elementary grammar-  writer is concerned with : he may suit his own  convenience, and contend the point as he -I  pleases. Thus much is certain, and is quite  sufficient for our purpose, — that to Marcus ,  cannot be considered a dative case, nor he wiU ]  love a future tense, on any other principle than  the one it is stated to flow from, namely;  that marked i. in Sect. 21.   23. To the practical grammarian we may     64     ON GnAMMAR.     [[chap. I.     likewise frequently allow, for the sake of con-  venience, the continuing a word under its  usual denomination, when its office, and con-  sequently its character, are essentially changed.  He will love, taking the three words as one  expression, are a verb both on the principles  we have ascertained, and in the practice of  the elementary grammarian : but in parsing  tliis verb — this p^iio, dictum, communication,  01 sentence, — only one of the three words  can properly retain the denomination of verb,  viz. that word to which the others have a re-  ference, by which they hang together, and  are signified to be a sentence, namely, ■will.  As to the word love, which the practical  grammarian will tell us is a verb in the infi-  nitive mood, it does not in fact fulfil the  office of a verb, but of a substantive. But if,  by calling it a verb in the infinitive mood, its  character for practical purposes is con-  veniently marked, we may fairly leave the  matter as it stands. All we insist upon is,  that the doubtful character of the word is a     SECT. 23.] ON GBAMMAB. 63   consequence of the principle marked ii. in  Sect 21."     I • Strictly, there is no verb but when a c  cation ib actually made ; and that word is then the  verb, which expreaseB the communicatioti, or which,  when several words are necessary, ie the sign of union  among the whole of them. A verb not actually in  use is acaptain out of commission, and if we still call  it a verb, it is by courtesy. Home Tooke never an-  swered his own question, " What is that peculiar dif-  ferential circumstance, which added to the definition  of a noun, constitutes a verb ?" (Diversions of Purley,  Vol. II. p. 514),) because he bad previously blinded  himself to the perception of what it is, by laying down  the principle already animadverted upon in a note ap^  ponded to Sect. 3., namely, that the business of the  mind, as far as regards language, extends no fiirther  than to receive impressions: the consequence of which  priuciple would be, (if it could have any consequence  at all,) that the first invented elements of speech were  nouns, or names for those impressions ; which accord-  ingly seems to be his notion, and that verba afterwards  arose from nouns, by assuming the difierential some^  thing that was found to be wanting. Our doctrine is,  that the original element of speech contained both the  artificial noun and the artiiicial verb ; that the mind  exerted its active powers in order to evolve the artir  ficial parts ; that the act of joining them together     M ON GRAMMAR. [CHAP. I.   S't. It might also perhaps admit of dis-  pute, whether substantives in what are called  their oblique cases, do not, by being the ad-  juncts to other words, and taking a change  of form to signify their servitude, cease in  fact to be substantives, and merit no higher  name than adjectives or adverbs. But here  again we consult convenience by using the  descriptive title, a substantive in the geni-  tive, dative, accusative, or ablative case. We  only need insist, as philosophical inquirers,  that the definition of a substantive in Sect.  15., is not less correct, because it does not in-  clude a substantive in these oblique cases*.   i^ain, made them a verb ; but if the title was given to  one more than to the other, it was given to that which  arose most immediately from the occasion, and took  the other to fis or determine it ; and that subsequently  that word in a sentence came to be coneidcred the  verb, which joined the parts K^ether, and signified  them to be a sentence.   * The only oblique case in English substantives,  is the genitive terminating in 'fi or having only the  apostrophe, the s being elided. Grammarians, in-  deed, have found it necessary to allow an accusative.     SECT. 25.] ON GRAMMAR. G5   25. The very doubt itself which so often  arises, whether a word is this or that part of  speech, — the varying classification of the parts  of speech by different grammarians, — are cir-  cumstances entirely favourable to the theory  advanced, and adverse to any theory which  attempts to explain the parts of speech by a  reference to the nature of our thoughts in-  dependently of language. For if the parts of  speech had taken their origin from this cause*   because pronouns have it : for if  in the sentence Cas-  s-iua loved him, we put the noun where the pronoun  stands, and say, Casmus loved Brutus, it seems con-  venient to consider the noun to be in the same case  that the pronoun was in. On the same principle, the  substantives which, in the classical languages, have no  accusative distinct from the nominative, are neverthe-  less considered to have an accusative, because, lite  other substantives, they can be used objectively with  regard to verbs active and certain prepositions. On  the score of convenienee this must be allowed. But  when words are taken separately, (and this, by  the very delinttion of the word, is the business of  parsing,) it is evident that only those substantives  are, strictly speaking, in the accusative case, which,  when uaed as just staled, have a form to signify it.     ON GRAMMAR.     [chap.     surely we could never have been in doubt  either as to vskat, or koio many, they were.  But our theory accounts at once for the in-  certitude on these, and many other points.  We admit no original element of speech but  the VERB, or that one sign which denotes what  the speaker wishes to communicate. If no  one sign can be found adequate to the occa-  sion, then we must make up a sign out of two  or more. Now the division of a verb into  these parts of speech, is necessarily attended  by the consequence, that each part is insigni-  ficant of a communication by itself, and that  they signify it only by being joined together.  Supposing a sentence never consisted but of  two parts, the mere act of joining them to-  gether, would be sufficient to signify that  they were a sentence or verb. But the ne-  cessity or usage of speech being such, that  the hearer knows a sentence may consist of  two or of many words, how is he to be warned  that a sentence is formed, unless to certain  words is given the power of signifying a sen-     SECT. 25.] ON GRAMMAR. G?   tence, while to other words this power is de-  nied until associated with a word of the for-  mer class? Hence the distinction between  noun and verb ; a distinction arising out of  the necessities of speech, and not out of the  nature of our thoughts. The noun and the  verb, then, are the original parts of speech, the  verb beingthepreviouselementof both. But  as each derives its office and character solely  from an understanding between the speaker  and the hearer, a change of understanding  may make them change their offices, and so  the verb shall sometimes be a noun, and the  noun a verb. These changes occur in fact  so frequently, as to require no example.  Then, as we have seen, a noun will frequently  be used as the adjunct of another noun, and  so become an adjective j an adjective or other  word may be joined to a verb, and so become  an adverb j and any of these, by frequent use  in particular combinations, may acquire, or  seem to acquire, a new and peculiar office,  and so become articles, prepositions, and con-     ON GRAMMAR.     [chap. I.     junctions. But who can ascertain that de-  gree of use, which, to the satisfaction of every  grammarian, shall fix them in their acquired  character • ? Nay, must not every such word,  of necessity, while in transitu, be at one period  quite uncertain in its character ? In this man-  ner do the effects arising out of such a theory of  the parts of speech as we have supposed, agree  with actual effects, and fully explain them.  26. Again, on any other hypothesis than  the one before us, what are we to think of  compounded nouns, adjectives, verbs, adverbs,  &c., of which all languages are full ? With-  out adverting to established compounds, such  as (to take the first that occur) husbandman.     * What, for instance, shnll we call the word fi/ce in  such phrases as like him, like me? Originally theword  unto intervening between it and the pronoun, govern-  ed the latter ; but unio cannot now be aid to govern  the pronoun, since it has been so long disused, as to be  no longer mtderstood. We miglit therefore say, that  like is a preposition governing the pronoun : — the  point perhaps is disputed ; — be it so : for this fact jugt  serves our argument.     :6.]     ON ORAMMAK.     m     worJcmanlike, waylay, browbeat, nevertheless ;  without bringing words from the ilUmitably  compounded Greek language, — we may refer  to such as are not established, but compounded  ibr the particular purpose ; as when Locke  speaksof '* Mr. 'Nev/ton'sjiever-enough-io be ad-  mired book," where the words in italic are an  adjective; and when some old lady pettishly  says to her grandchild " Don't dear Grand'  mother me i" v/here the whole sentence, ex-  cept the pronoun governed in the accusative,  is a verb. So in the phrases to fiAxov <rvvoia-eiv  7^ iroXei the being-about-to-be'prqfitable-to-t/ie-   Ci'/y,— and, TO Tct Tou iroXefiov raj^ii xal Kara   Kaipov Trpa.TTea$at, the completing-spcedili/'and-  seasonablif-the'lhings-for-the-war, we are war-  ranted in considering the whole of the words  following the article, to be, in each instance,  a noun-substantive. For these, and for every  other species of compound, the theory before  US at once accounts. For it shows that the  use of many words to form one sentence, arises  out of the necessities of language only, the na-     w     ON GRAUMAR.     [^CHAP. I     tiira] impulse of the mind being tomake its com-  munication by a single expression. Having  complied, then, with the necessities of lan-  guage, and rendered it capable of serving as  the interpreter of much more knowledge than  we could have attained without its help ; we  then return on our steps, and give a unity to  our expressions in every possible way.   27. The corruption of early phrases, by  which, in so many instances, they come under  the denomination of adverb, will be found  another obvious consequence of the present  theory, while they abundantly perplex the  grammarian who attempts to reconcile them to  any other system. "Omnis pars orationis"  says Servius, "quando desinit esse quod est,  migrat in adverbium." " I think" says Home  Tooke, " I can translate this intelligibly —  Every word, quando desinit esse quod est,  when a grammarian knows not what to make  of it, migrat in adverbium, he calls an ad-  verb."* What indeed can be made of such     ' Divctsioiia vi Puiky, Vol. I.     SECT. 270 Of* GRAMMAR.     71     expressions as at all, by and by, to be sure, for  ever, long ago, no, yes. They are adverbs,  say the grammarians. But (to take the  phrases first) what are the words, individually,  of which the adverbs are composed? The  answer will be, they are prepositions, adjec-  tives, &c., which remain from the corruption  of regular phrases once in use. This is a true ,  account of the matter : — yet it leaves us still  to ask, what ai'e these single words, now that  the phrases which produced them exist no  longer in their original state. Let any gram-  marian, if he can, prove their right to the  name of any of the received parts of speech.  Our system, if it does not make a provision  tor them by a name for a new class of words,  at least shows the cause and the nature of their  difference. For according to our principles,  words have both a separate and a, joint signifi-  cation. But if words should be constantly     another place, he says " that this class of words, (ad-  verb,) is the common sink and repository of all hetero-  geneous, unknown corruptions."     w     ON GRAMMAR.     [chap, r.     occurring in particular combination, this ef-  fect will enaue, — that their separate significa-  tion in such hackneyed phrase, will at last be  quite unattended to, and their joint significa-  tion alone regarded ; — and such phrases will  then be as liable to be clipped in the currency  of speech, as any long word which is trouble-  some to be uttered at full : — thus will the re-  maining parts of the phrase be fixed for ever  in their joint, and lose for ever their separate  signification*. So much for the words com-  posing adverbial phrases. But what are we  to say for no, yes, which probably had the  same origin as the phrases ? These have not,  Hke the phrases, a compound form, nor do  they, like the phrases, always assist in making  up a sentence, but are frequently and proper-  ly pointed oft' by the full stop. Are we, un-  der such circumstances, to call them adverbs P  •• Yes." This is the answer our grammarians  make. But is there, in these words, any     • Thcwordtoas asignofthcinfiiiitivL'moodcumcs  onilcr this doicnption.     SECT. 28.] ON GRAMMAR.     73     thing which gives them a just claim to be  ranked with any of the received classes of  words? " No." This is an assertion it would  be difficult to gainsay. For consider them  well, and we shall find, that, in their present  use, they are not j3ar/s of speech at all, except  with reference to the larger portions of dis-  course of which all the sentences are parts :  they are sentences ; and they afford a striking  example of what was intimated in the prece-  ding section, namely the tendency oflanguage,  in a mature state, to return on its early steps  as far as can be done without losing the ad-  vantages gained : for not only do we, when-  ever we can, bring the smaller parts of speech  into such union as to form larger parts, but  in some instances, (as in these last,) we come  round again to the simpHcity of natural signs.  28. This union of the smaller into larger  parts of speech, and the power we have to dis-  pose the same materials into more or fewer  sentences, will furnish further proofs, that the  present theory of language can alone be the     74     ON GRAMMAR.     [chap.     true one. A proper examination of compound  sentences will show, that the grammatical  parts into which they are first resolvable, are  not the single words, but the clauses which  are formed by those words ; which clauses are  substantives, and verbs, and adjectives, and  adverbs, with respect to the whole sentence,  however they may, in their turn, be resolva-  ble into subordinate parts of speech bearing  the same or other names. To take the fol-  lowing as an example : " The sun which set  this evening in the west, will rise tomorrow  morning in the east." The two parts into  which this sentence is resolvable, are, to all  intents and purposes, a noun-substantive and  a verb, if considered with respect to the whole  sentence*. This is the first, or broadest ana-   * And HO may the two parts (technically called the  protasis and apodosis) of every periodic sentence be  considered : for every period, (TEfi'ofos, a circle,) is re-  solvable into two chief parts, the one assimilated to  the semicircle tending out, the other to the rendering-  in, or completing semicircle. These answering parts  ate commonly indicated in Greek by iJth — ft; in En-     *■]     ON CRAMMAlt.     75     lysis. Then taking the former of these two  chief constructive parts, we shall find it re-  solvable into these two subordinate parts, viz.  the sun, a noun substantive, and w?iick set this  evening in the west, its adjunct or adjective : —  the latter chief constructive part being in the  same way resolvable into will rise, a verb, —  and, tomorrow morning in the east, its ad-  junct or adverb. Returning to the adjective  of the former chief constructive part, we shall   gUsh very frequently by as — so; though — yet, &c.  There may exist a doubt in most sentences so construct-  ed, whether the one part has a claim to be considered  tlie verb more than the other : each part is meant to be  insignificant by itself, and, {as was lately supposed of  the parts of speech in their early institution, before a  sentence was composed of more than two words,) they  Bifrnify a communication by the very act of being join-  ed together. Yet as the protasis is a clause in sus-  pense, and so resembles a substantive in the nomina-  tive case before the verb is enounced ; — as the apodo-  618 removes the suspense, and so resembles the verb in  its effect on tlie substantive ; — it seems that in con-  Hidering the protasis as a nominative case and the apo-  dosis aa its verb, we shall not be far from taking a ,  right view of the principle and procedure.     76     ON GRAMMAR.     [chap.     find it, if separately viewed, to be a sentence  having its nominative which, its verb set, and  the latter having its adverb tins evening in the  ivest ; which adverb is resolvable into two  clauses of which the former consists of the de-  monstrative adjective this, and evening, a sub-  stantive used objectively with relation to the  preposition on understood •• The latter clause  in the west is nearly similar in its grammatical  parts ; but the preposition it depends upon, is  not understood. This subordinate or adjec-  tived sentence which we have thus taken to  pieces, (viz. which set this evening in the west,')  is however no sentence when considered with   " Or more properly this eeening is an adverb ; for a  word cannot justly be called understood, when its ab-  sence is not suspected till the grammarian informg us of  it : — on before euch phrases when the custom to omit  it had just begun, was indeed understood; it is now  understood no longer, and what remains of any such  phrase is an adverb. As the next clauses, in the tceat,  retains its preposition, we are at liberty to parse the  clause, instead of considering it, in the whole, as an  adverb attcndijig the verb set, though we are also ab  liberty to consider it in the latter way.      *■]     ON GilAMMAR.     77     reference to the larger sentence of which it is  a grammatical part : but it might, if the  speaker had pleased, have been kept distinct,  and the same meaning have been conveyed by  two simple sentences, as by the one com-  pound one : e. g. " The sun set this evening  in the west : — It will rise tomorrow morning  in the east." Here, we have two sentences or  commuuications. But this is nothing more  than a difference in the manner of conveying  the thought, precisely analogous to the using  of two words that restrict each other, in place  of a single appropriate sign. In the instance  before us, the thought, whether expressed by  the one sentence or the two, is the same ; and  it is one and entire, whatever the expression  may be. For we must not confound the two  facts referred to in the sentences, with what  the mind thinks of the facts : — it is the con-  nexion of the facts that the speaker seeks to  make known. Yet he may imagine he can  best make it known by using the two sen-  tences ; for though, it is true, that while they     78     ON GRAMMAR.     [chap. I.     are in progress, they will be understood se-  parately, yet no sooner will they be com.  pleted, than the hearer will understand them  limited and determined the one by the other,  and no longer abstractedly as while they were  in progress. In this manner, in correspond-  ence with the principle stated Sect. 21 . iii., will  the same result be obtained by the two, as by  tlie one sentence.   29. This power, which exists in all lan-  guages, of expressing the same thought in a  variety of different ways, is, one would think,  a suiEcient proof, by itself; that thoughts and  words have not the kind of correspondence  whicli is commonly imagined : for if such cor-  respondence had existed, the same thoughts  would always have been expressed, if not by  the same words, yet by words of similar mean-  ing in the same order. Let us suppose that  tlie expressing a thought by several words,'  I had been, (which it is not,) a process analo-  gous to that of expressing the combined sounds  of a single word by several letters. There is     SECT. 29.] ON GRAMMAR.     79     the more propriety in instituting tlie compa-  rison, because men were driven to the latter  expedient by a necessity similar to that which  drove them to the former. For, no doubt,  the first idea of the inventors of writing was,  to appropriate a character for every word ; and  we are told that, to this day, a practice near  to this prevails in China, But it was soon  found that the immense number of characters  this would require, must make the completion  of the design next to impracticable ; and the  expedient was at length adopted of spelling  words. By this expedient, twenty four cha-  racters, by their endless varieties of position  with each other, are capable of signifying the  multitude of words, and the innumerable sen-  tences, which constitute speech. The parts  of speech were set on foot by a similar urgency,  and in tlie same way. At first, every sound  was a sentence. But the communications  which the business of life required, far, far  outnumbered every possible variety of sound.  It was fortunate, therefore, when a necessity     eo     ON CnAMMAR.     [chap. I,     arose to give to some of the sounds a less par-  ticular application ; for then the requisite sign  was formed out of two or more sounds already  in use, and no new sound was required. So  far the parallel holds ; but it will go no further.  In the spelling of words by letters, the same  letters must always be used, — if not the same  characters, yet characters of the same power.  And it would have been the same in spelling  a thought by words, if the process had been  what it is commonly supposed to be :— that is  to say, the same thought would always have  been expressed by the same words, or if  the words had been changed, the change  must have been word for word, as in a  completely literal translation from one lan-  guage to another. How different this is  from fact, hardly needs further examples in  proof. Mr. Harris attempts to shew *, that     • Hermes, Book I. Chap. 8. We cordially agree  in Home Tooke's opinion of thia well-known work,  that it is " an improved compilation of almost all the  enors which grammarians liave been accumulating     SECT. 290 *^^ GRAMMAR.     81     tlic different forms or modes of sentences,  depend on the nature of our thoughts. That  the character of a thought has an influence in  determming our preference of this or that  mode of speech, needs not be questioned; but  all the modes of speech, are interchangeable  at pleasure, and therefore they cannot aub-  stantiallydepend on thenature of our thoughts.  An affirmative sentence, " 1 am going out of  town," ma be made imperative, " know,  that I am going out of town ;" or interrogative,  *' Is it necessary to say, that I am going out  of town ?" A negative sentence, " No man is  immortal," maybe made affirmative, "Every  man is mortal." It would waste time and  patience to multiply examples. The con-  clusion, then, is, that the parts of speech and   from the time of Aristotle, to our present days." Di-  versions of Furley, Vol. I. page 120. Vet occasionally,  when our etymologist runs a little bard on this Com-  piler of errors, the theory we advance, opposite as it ib  in its general tenor to all that the Hermes conttuns,  will be found to lend its author a lift. See the section  ensuing in the text.     ON CnAMMA     [chap.     the forms of sentences, are alike attributable  to the necessities and conveniences of lan-  guage, and not to the nature of our thoughts  independently of language. Perhaps by this  time it may almost seem that an opinion con-  trary to this has no defined existence, and that  the combat has been against a shadow. But  this is not true. If the opinion opposed to the  principles contended for, is seldom ^rwio%  expressed, it is nevertheless universally under-  stood — it is at the bottom of all the systems  of grammar, of logic, and of rhetoric, which  we study in our youth, and which we after-  wards make our children study ; and as it is  an opinion radically, essentially wrong, the  pains employed to overthrow it, cannot, if  successful, have been supeiHuous. In no  other way was a preparation to be made for  an outline of the higher departments of Sema-  tology.   30. New, however, as we believe our  theory to be, yet it is not without authorities in  its favour ; and with these we shall conclude     SECT. 30.]     ON GRAMMAR.     the chapter. Harris, the author of" Hermes,"  in treating of connectives, stumbles unawares  on the fact, that a word which is significant  when alone, may he no significant part of  what is meant hy the expression it helps to  form. He makes nothing indeed of the fact,  further than to lay himself open to the ridicule  of Home Tooke for tKe inconsistent assertions  in which it involves him. " Having " says  Tooke *, " defined a word to he a sound sig-  nificant, he (viz. Harris) now defines a pre-  position to be a word devoid of signification ;  and a few pages after, he says, ' prepositions  commonly transfuse something of their own  meaning into the words with which they are  compounded.' Now if I agree with him,"  continues Tooke, " that words ai'e sounds  significant, how can I agree that there are  sorts of words devoid of signification ? And if  I could suppose that prepositions are devoid  of signification, how could I afterwards allow,     ' Diversions of Purley, Vol. I. Cliap. 9.     9»     ON GRAMMAR.     [chap. T.     that they transfuse something of their own  meaning?" Yet with all this, Harris is right,  only that he is not aware of the principle,  which lies at the bottom of his own doctriue.  A preposition, as well as every other word,  is a sound significant j — it has an independent  abstract signification : but being joined into a  sentence, it is devoid of that signification it  had when alone : it has then transfused its  own meaning into the word with which It is  compounded, as that word has transfused its  meaning into the preposition — that is to say,  they have but one meaning between them.   31. But Dugaid Stewart, in his Philoso-  phical Essays, furnishes a direct, and a more  satisfactory authority in favour of the theory  we have advanced. " In reading " says he •,  " the enunciation of a preposition, we are apt  to fancy, that for every word contained in it,  there is an idea presented to the understand-  ing ; from the combination and comparison of  which ideas, results that act of the mind  • Philosophical Essays, Essay 5. Chap. I.     SECT. 31.] ON GRAMMAR, 85   called judgment. So different is all this from  fact, that our words, when examined sepa-  rately, are often as completely insignificant aa  the letters of which they are composed, de-  riving their meaning solely from the connexion  or relation in which they stand to others." —  Again : " When we listen to a language which  admits of such transpositions in the arrange-  ment of words as are familiar to us in Latin,  the artificial structure of the discourse  suspends, in a great measure, our conjectures  about the sense, till, at the close of the  period, the verb, in the very instant of its  utterance, unriddles the jenigma. Previous  to this, the former words and phrases resemble  those detached and unmeaning patches of  different colours, which compose what op-  ticians call an anamorphosis ; while the effect  of the verb, at the end, may be compared to  that of the mirror, by which the anamorphosis  is reformed, and which combines these appa-  rently fortuitous materials, into a beautiful  portrait or landscape. In instances of this     k      86 ON GRAMMAR. j^CHAP. I.   sort, it will generally be found, upon an  accurate examination, that the intellectual  act, as far as we are able to trace it, is  altogether simple, and incapable of analysis ;  and that the elements into which we flatter  ourselves we have resolved it, are nothing  more than the grammatical elements of  speech j — the logical doctrine about the com-  parison of ideas, bearing a much closer  affinity to the task of a school-boy in parsing  his lesson, than to the researches of philoso-  phers able to form a just conception of the  mystery to be explained." — Had this acute  philosopher brought these views of language  to the elucidation of Grammar, Logic, and  Rhetoric, and so have cleared them from the  incrusted errors of immemorial antiquity,  the reader's patience would not have been  tried by the chapter now finished and those  which are to follow.      CHAPTER 11.     ON LOGIC.     Say, first, of God above, or man below.   What can we reason, but from what we know.   POPE.   1. In commencing this branch of Semato-  logy, it may be as well to define not only this  but the other branches, that their presumed  relation and difference may at once appear :   i. Grammar, then, is the right use of  words with a view to their several functions  and inflexions in forming them into sentences ;   ii. Logic is the right use of words with a  view to the investigation of truth ; and   iii. Rhetoric is the right use of words with  a view to inform, convince, or persuade *.   * This definition includes the poet^s use of words  as well as that of every other person, who, having one  or more of the purposes mentioned in view, speaks or     fts     ON LOGIC.     [chap. II.     2, The object of the present chapter  will be, to show that there is no art of Logic  (except sucli as is an imposition on the un-  derstanding but that which arises out of the  principles ascertained in the previous chap-  ter ; — that tliis, which is the Logic every man  uses, agrees with the definition in the previ-  ous section; —and that we cannot carry the  definition further, without transgressing a  clearly marked line which will usefidly distin-  guish between Logic and Rhetoric.   3. In affirming that there is no art of Lo-  gic but that which arises out of the use of  signs, we do not mean that reason itself is de-  writes skilfully. Should it be said, that the poet's end  is to delight, — we answer that he gains this end by in-  forming, convincing, or persuading. The true dis-  tinction between the poet and any other speaker or wri-  ter, lies iu the different nature of their thoughts, In  communicating his thoughts, the poet, like others who  are skilful in the use of words to inform, convince, or  persuade, is a rhetorician ; although, with reference to  the creative genius displayed, {iroix^n a jrcn'm,) and al-  so with reference to the added ornament of metre or  rhyme, we chU the result, a poem.     SECT. 4.3     89     pendent on language. Reason must exist pri-  or to language, or language could not be in-  Vented or adopted. What we affirm is, that  prior to the use of words or equivalent signs,  »o art exists : the mind then perceives, as far  fts its powers extend, intuitively; and thus  working without media, it can no morye ope-  rate otherwise than as at first, than the eye  can see otherwise than nature enables it. The  mind can, however, invent the means to assist  its operations, as it has invented the telescope  to assist the eye ; the difference being, that  the telescope is not such an instrument as all  minds would invent, but the use of signs to  assist its operations, grows out of the human  mind by its very constitution, and the influ-  ence of society upon that constitution.   4. That writers on Logic do not in gene- '  ral view the matter in this light, is evident  from this, that they devote, or at least they  persuade themselves and their readers that  they devote, a great pait of their considera-  tion to the operations of the mind indepeud-     90     ON LOGIC.     [chap. II.     entlyof language, which, for any practical end,  must evidently be nugatory on the supposi-  tion stated above ; since, if the mind, without  the aid of signs, can but operate as nature en-  ables it, all instruction concerning what the  mind does by itself*, will but be an attempt   * WattB Bays t&at " the design of Logic, b to  teaeli us the right use of our reason." Recurring to   our comparisDU in the previous section, this is as if any  one had proposed to teach the right use of the eye. It  is true indeed, a man may be taught a right use of the  eye, — that is, he may be taught to observe proper ob-  jects by its means ; and so may he be taught a right  use of reason by applying it to those things which are  conducive to his improvement and happiness. But all  this belongs to Morals not to Logic ; nor was this  Watts's meaning. He imagined a man could be tattght  how to use his reason independently of any considera-  tion of an instrument to work with ; as if any one had  offered to teach mankind how to sec with their eyes.  Now, there is nothing preposterous in offering to show  how a telescope is to be used in order to assist the  eye ; nor any thing preposterous in trying to show  how words may be used in a better manner than com-  mon custom instructs us, in order to assist the  mind. — Be it observed that the objection here made,  is to what was proposed to be done by Watts, and not     SECT. 4.] ON LOGIC. 91   to teach us that which every one does with-  out teaching, and which no teaching can  make us do better : but if, by the use of signs,  the mind can carry its natural operations to  things which it could not reach without signs,  the instruction of the logician should at once  begin by pointing out the use and the abuse  of signs. Now this is in fact the point at  which every teacher of logic does begin, how-  ever he may disguise the real proceeding from  himself, and whatever confusion he may throw  over his subject, by not knowing in what way  he is concerned with it. In pretending to  teach us the nature of ideas j logicians do no-  thing but teach us what knowledge we attain   to what he actually does, except so far as he has done  it amiss from setting out badly. What follows in the  text will explain this last observation.   Our illustration must not lead the reader to think  we are ignorant of the fact that men do learn to see,  that is, to correct, by experience and judgment, the im-  pression of objects on the retina. We take the matter  as commonly understood, namely, that men see correct-  ly by nature, which is near enough to the truth for our  present purpose.     IB ON LOGIC. QCHAP. II.   by means of words-, and when Home Tooke  says of Locke's great work, that it is " merely  a grammatical Essay or Treatise on words," *  be comes so near the truth, that it is wonder-  ful he should have so wrongly interpreted  other parts of that philosopher's doctrine.  Putting a wrong construction on Locke's just  fundamental principle, that the mind has no  innate ideas, Tooke affirms that '* the busi-  ness of the mind, as far as it regards language,  extends no further than to receive impres-  sions, that is, to have sensations or feelings.  What are called its operations are merely the  operations of language." t This is palpably  absurd ; ftx how can language operate of it-     • Diversions of I'utley, Vol. I. page 31, note.   -j- Diversions of Purley, Vol. I. page 51. We have  already quoted this passage ; and perhaps more than  ontc : but it is hoped we need not apologise for the re-  petitions whicli may be found in this and the next  chapter. Our purpose is to trace Grammar, Logic, and  Rhetoric, to a common source, and in doing so, if they  really have an origin in common, we must necesEarily  traverse the same ground repeatedly to come at it     SECT. 4.]     93     aelf? The mind must observe, compare, and  judge *, before it can invent or adopt the lan-  guage of art ; and having adopted it, every  use of it is an exercise of the reasoning facul-  ty, excepting only that kind of instinctive use,  in which some short sentence takes the place  of a natural ejaculation. Feelings or sensa-  tions we cannot help having ; but these do not  help us to language. This requires the ac-  tive powers of the mind ; and every word, in-  dividually, will accordingly be found the sign  of something we kno-w, obtained, as every  thing we know must be obtained, by previous  acts of comparison and judgment, involving,   * These powers of the mind are innate, — that is  to e&y, they belong to tlie mind by its constitution, al-  though sensation is the appointed means for first call-  ing them forth. It should seem as if Tooke thought  nothing was bom with man except the power to receive  senEStionB or feelings, and that reason comes from Un-  guage ; an opinion so preposterous that we can hardly  think him capable of it ; and yet, from what he says,  no other can be understood : — " Jleason,"" he says, " ia  the result of the senses, and of experience." Diver-  sions of Purley, Vol. 11, p^e 16.     J^     ON LOGIC.     [CilAP. II,     in every instance beyond that which sets the  sign on foot, an inference gained by the  use of a medium. And such, as we have seen,  are the necessities of speech, that tliey lead  us constantly to extend the application of  words ; which extension requires new acts of  comparison and judgment; and thus, by  means of words, (or signs equivalent to words,)  we are constantly adding to our knowledge,  still carrying the signs with us, to mark and  contain it, and to serve afterwards as the media  for reaching new conclusions. It is only ne-  cessary to read Locke's Essay with this ac-  count of the matter in view, to prove that it  is the true account j so readily will all that he  has said on ideas, yield to this simple inter-  pretation *, He who first made use of words     * " Read," saya Home Tookc, " the Essay on the  Underslnnding over with attention, and see whether  all that its immortal author has justly concluded, will  not hold equally true and clear, if we substitute the  composition, &c. of lerraa, wherever he has supposed a  composition, Sec. of ideas. And if that, upon strict  examination, appear to you to be the case, you will     SECT. 4.]     95     equivalent to yellow, white, heat, cold, soft,  hard, bitter, sweet*, used them, respectivelyy  to signify the individual sensation he was con-  scious of, and in that first use, the expression  must have been a sentence, or tantamount to  a sentence. By experience, he came to know  the exterior cause of that sensation, and after-  wards, by the same means, to know that other  need no other argument against the composition of  ideas : it being exactly similar to that unanswerable one  which Mr. Locke himself declares to be sufficient  against their being innate. For the supposition is un-  necessary : every purpose for which the composition  of ideas was imagined being more easily and naturally  answered by the composition of terms, whilst at the  same time it does likewise clear up many difficulties in  which the supposed composition of ideas necessarily in-  volves us." Diversions of Purley, Vol, I. page 38.  In this, and other passages, H. Tooke is very near  the trutli ; but he nevertheless misses it. " The com-  position, Sic. of terms "' in lieu of " the composition, &c.  of ideas," does not describe the actual process. But  Tooke, who discovers that Locke has started at a  wrong place, begins his own theory from a false found-4  ation.   • yide Locke, B. 2. ad initium : we have used  the examples before. Chap. I, Sect. 16.     ».     ON LOGIC.     [chap. It,     ol^ects produced the same sensation. To  these several objects he would naturally apply  the expression (originally tantamount to a sen-  tence) by which he first signified the sensa-  tion ; and suppose those objects already pro-  vided with namesj the expression would, in  such pew application, be tantamount to a  name or noun-adjective. Thus in the several  instances, he would use two names for one  thing, in correspondence with our present  practice when we say, yclhw flower, yellow  sky, yellow earth, yellow skin. Such a proce-  dure is an effect and a proof of what the speak-  er has observed in common, and of what he  observes to be different, in the several ob-  jects; and this is a knowledge evidently ob-  tained from comparison and judgment exer-  cised on many particulars. The same know-  ledge enables us, when we please, to drop the  words which name the objects accojding to  their differences, and to retain only that which  signifies their similarity, and the name-adjec-  tiv e then becomes a name-substantive stand-     gECT, 5.3 ON LOGIC. 97   ing for the sensation itself whenever or how4  ever produced, and not standing for it in amy  particular case, until limited to do so by the  assistance of other words. Individually and  separately, then, these words^ viz. yellow;  white, heat, cold, soft, &c. are, to him who  has properly used them in particulars, tiie  eigns of the knowledge he ha^ gained by com^  paring those particulars :«^hey denote con-  clusions arising out of a rational process which  has been carried on by their means ; which  conclusion, as to the word^elloWf for instaop^  is this, — ^that there are » great mwy Qbjepte  which produce the same sensation, or a sensar  tion very nearly the same j*— ^(very nearly the  same, since yeU&w^ by all who have acquired  a full use of the word, is applied to different  shades of yellow j — ) and to understand the  word, is to have arrived at, or kno^ this cof^-  elusion.   5. The words so far referred to, are those  which denote what Locke calls simple ide^js.  Now, we may reasonably doubt wheth^ the   H     98 ON LOGIC. ^CHAP. II.   mind could have obtained the knowledge,  which, as we have seen, is included even in  a word of this kind, if it had not been gifted  with the power of inventing a sign to assist  itself in the operation. That sign needs not  be a word, though words are the signs com-  monly used. He who remembers the sensa-  tion of colour produced by a crocus, is re-  minded of the crocus the next time he has  the same sensation from a different thing ;  and the crocus may become the sign of that  sensation arising from the new object, and  from every future one. And this is the way  in which the mind probably assists itself an-  tecedently to the use of language, or where,  (as in the case of the totally deaf *,) the use of     * Though long for a quotation, yet we cannot re-  sist transcribing, from a work by Dr. Watson, master  of the Deaf and Dumb Asylum, Kent Road, near  London, the following able remarks : — they will help  to shew how for superior are audible signs to every  other kind, and place in its proper light the misfor-  tune of being naturally incapable of them. He is  speaking of the comparative importance of the two     SECT. 5.3     99     it, by the ordinary means of attainment, is  precluded. But for this power of the mind,   senBES, hearing and seeing. " Were the point," he  says, " to be determined by the value of the direct  sensations transmitted to the sensorium through each  of them, merely as direct sensations, there could not  be any ground for a moment's hesitation in pro. ,  nouncing the almost infinite superiority of the ej/e to ]  the ear. For what is the sum of that which we derive I  from the car as direct sensation P It is sound ; and  sound indeed admits of infinite variety ; but strip it of j  the value it derives Irom arbitrary associations, and it  is but a titillation of the organ of sense, painful or  pleasurable according as it is shrilly soft, rough, dis-  cordant, or harmonious, Sec. Should one, on tlic con-  trary, attempt to set forth the sum of the information we  derive from the eye " — independently of the aid derived  from arbitrary means — " it is so immense, that volumes  could not contain a full description of it ; so precious,  ' that no words short of those we apply to the mind itself,  can adequately express its value. Indeed, all lan-  guages bear witness to this, by figuratively adopting  visible imagery to signify the highest operations of in-  tellect. Expunge such imagery from any language,  and what will be left ! What, in this case, must be-  come of the most admired productions of human ge-  nius P Whence then (and the question is often asked) 1  does it arise, that those bom blind have such su-  h2     100 ON LOGIC. [chap. 11.   which seems pecuHai* to man, and is the  cause of language, (not the effect of it, as     perlority of imelligence over those bom deaf? Take,  it miglit be said, ii boy nine or ten years of age who  has never seen the light, and you will find him con-  versable, and ready to give long narratives of past oc-  currenceH, &c. Place by his side a boy of the same  age who baa had the misfortune to be bom deaf, and  observe the contrast. The latter is insensible to all  you say : he smiles, perhaps, and his countenance ie  brightened by tlie beams of ' holy light;' he enjoys  the face of nature; nay, reads with attention your  features ; and, by sympathy, reflects your smile or  your frown. But he remains mute : he gives no ac-  count of past experience or of future hopes. You at-  tempt to draw something of this sort from him : he  tries to understand, and to make himself understood ;  but he cannot. He becomes embarrassed : you feci  for him, and turn away from a scene so trying,  under an impression that, of these two children of mi^  fortune, the com])ari8on is greatly in favour of the  blind, who appears, by his language, to enter into all  your feelings and conceptions, while the unfortunate  deaf mute can hardly be regarded as a rational  being ; yet he possesses all the advantages of vi-  sual information. All this is true. But the cause  of this apparent superiority of intelligence in the blind,  is seldom properly understood. It is not that those     SECT. 5.3 ON LOGIC. 10]   H. Tooke seems to tliiak,) we never should  have been able to arrange olyects in classes,   who are blind possess a greater, or anything like an  equai stock of materiak for mental op^adons, but bs-  cause they possess an invaluable etigine for forward-  ing those operotioiis, however scanty the materials to  operate upon — artificial language. Language is de-  fined to be the expression of thought ; so it is : but it  is, moreover, the medium of thinking. Its value U>  man is nearly equivalent to that of his reasoning fa-  culties: without it, he would hardly be rational. It  is the want of language, and not the want of hearing,  (unless as being the cause of the wont of language,)  that occasions that deficiency of intelligence or ine&.  pansion of the reasoning faculty, so observable in the  naturally deaf and dumb. Give them but language,  by which they may designate, compare, classiiy, an4  consequently remember, excite, and express their sen^  sations and ideas, — then they must surpass the origin<  ally and permanently blind in intellectual perspicuity  and correctness of comprehension, (as far as having  kctual ideas afiixed to words and phrases is concerned,)  by as much as the sense of seeing, furnishes matter for  mental operations beyond the sense of hearing, con-  Eidered as direct sensation. It is one thing to have a^  fluency of words, and quite another to have correct no-  tions or precise ideas annexed to them. But though  the car furnishes us only with the sensation of sound,     t^ ON LOGIC. [chap. II.   and reason on them when so arranged ; nor to  consider some common quality in many ob-  jects, separately from the objects themselves.  Every object might have produced the same  individual effect by the senses, which it now  produces, and have been recognized as the  same object when it produced the effect  again ; for all this happens to other animals,  as to man ; but to know a something in each  which is common to many, implies a remem-  brance of that something in the rest at the  time of perceiving each individually j and  how can this remembrance, (a remembrance   and sound, merely as such, can stand no comparlEOD  with the multiform, delightful, and important informa-  tion derived from visual imprestiioDS ; yet as sound  admits of such astonishing variety, (above all when  articulated,) and is associablc, at pleasure, in the mind  with our other sensations, and with our ideas," (notions,)  " it becomes the ready exponent or nomenclature of  thought ; and in this view is important indeed. It is  on thie account, chiefly, that the want of hearing is to  be deplored as a melancholy chasm in the human  frame.'" Instruction of the Deaf and Dumb, in 3 Vols.  Edit. 1809. Vol. I. p. 49.      SECT. 6.]     ON LOGIC.     103     not of the objects, but of a common some-  thing in all of them,) how can it be kept up,  but by a sign fitted to this duty ; which sign,  as just observed, may be either a word, or  one of the objects set up to denote the com-  mon characteristic, and retained in mind  Bolely for this purpose, in this representative  capacity ?   6. In proceeding from what are called by  Locke simple ideas to those he denominates [  complex, we shall find the account just given  equally applicable. The words he refers to .  under the threefold division of Modes, Sub-  stances, Relations, are, as our last examples,  signs of certain conclusions obtained from s  comparison of particulars. This is true even \  of a proper name ; for a proper name, as was '  shewn Chap. I. Sect. 3., does not denote an  individual as we actually perceive him, or as. J  we remember him at any one time ; but it J  denotes a notion, that is, a knowledge of him I  drawn out of, or separated from all our par- '     I04f oNr Lo&ic. [cHap. ii.   ticular perceptions *• For such an effect of  reason^ we have however nb certainty that  the superior powers of the huknan mind ar«  indispensable; nor is it eiisy to ascertaiq  any peculiar privilege it enjoys till we find  it rising from individuals to classes. As  soon as it sets up a sign to represent some  property, whether pure or mixed, which has  been observed iA many individuals,— or to re-   * It id aft efifect of reaisoiiing to know that a pa]>>  ticular act or situation, which enters into our percep-  tion or conception of an object, is not essential — to  know, for instance, tliat the act of walkiAg is ftot es-  iBentiAl to John. The reasoning by which «uch k^w-  ledge is acquired, occurs indeed so early, that the  operation is forgotten ; but there was a time when our  perceptions were without the knowledge, because they  had not been repeated i^ isu^ti^t hUtiibet to leHkbl^  the mind to make the BCcessary ootaipluidcms^ Th^  natives of the South Sea Islands^ when Cttptaia Cook  <8nd his companions first made their appearance among  them, took every sailor and his garments to be one  creature, and did not arrive at a different condhision,  but by o{>portuiiitte6 fdr comparicon.     5-]     ON LOGIC.     105     present the whole class of individuals, so  classed because of the common property, — ^it  displays a power of assisting itself which we  have no cause to think any of the inferior  animals enjoy. To ahew how this takes place  in producing what Locke calls complex ideas,  and which he subdivides into Modes, Sub-  stances, Relations, would only carry us onc^  more over the ground we have so often cur-   Lsorily traversed. We should have to shew,  for instance, how some word, at first equiva-  lent to a sentence, by which a man expressed  his delight at a particular visible object, came  to be a name for the object ; how this name  beauly, came to be applied as a noun-adjec-  tive to the nouns-subatantive of other objects  producing the same or a similar emotion j  how, by the continued application of this  noun-adjective, we kept on comparing innu?  merable particulars, till our knowledge (no-  tion) included a very wide class of things  very different indeed in other respects, — nay^  including objects of other senses than sight—     106     [chap. II.     but still, agreeing with each other in a certain  effect produced on the mind : and that then,  dropping the nouns-substantive of the nu-  merous individuals, we retained solely in con-  templation the noun beautiful or beauty, the  sign of the knowledge we had gained from  this extensive comparison— of the induction  derived from these numerous particulars *.   • Very few persons reach so wide a knowledge of  the subject as we here refer to, and books may be, and  have been written, to teach us how to apply the word  beautiful with taste, and critical — nay, moral pro-  priety. Having attained so far, we are not to suppose that  beautiful or beauty is a real existence independently  of the classification of objects we have thus established.  All we have learned is, to know the objects which pro-  duce a certain elfect ; to know why they produce it ; to  enjoy, it is probable, the pleasure of that effect with  higher relish ; and to be prepared, by means of the  classiUcation we have formed, to lise, in our reasonings  on the objects it contains, to higher truths, and still  more important conclusions. Now, if the reader would  see how a business so plain and simple, may appear  very complex and mysterious, let him consult  Plato  on the beautiful or t'o xayjtv, as he will find it treated,  for instance, in the dialogue called STMHOSION :  Let him admire as he will, (for who can help it.     SECT. 6-3 ON LOGIC. 107   We should again have to shew, (to take  another instance,) how a word once expres-  sive of some sentiment or recognition of  which a horse was the subject, came to be  used as a name for that particular horse i  that the name came afterwards to be given to  another resembling creature, — thence to  another, — and to others, till the points of re-  semblance which led to this extension of the  word, could be found no longer *. We should   especially in company with Cicero, — witness his Errare  tnekercule malo cum Plaione, quam cum istia vere  sentire?) let him admire the sublimity which the  amiable and highly-gifted Athenian throws over his  doctrine ; but let him not be betrayed into an opinion,  that a speculation which is in the most exalted etriun  liipoeh'y, belongs to the sober, the undazzled, and tin-  dazzling views of philosophy.   • Compare Chap. I.Sect, 10. We may be per-  mitted once more to observe, that, with regard to sab-  stances at least, the sign of the class needs not be a  word : one individual set up for all, will equally serve  the purpose. Not that the boundaries of a class are  plain, till an accurate logic determines them ; but the  general differences (as of the horse, for instance) are  sufficiently obvious to prevent a person from being     JflB     [chap. II.     likewise have toshew, (totake a third instance,)  how some word,-^originally equivalent, like  the others, to a sentence, — by which a man  expressed his gratitude for kind offices, might  come to be a name for every one to whom  gratitude for similar offices was due; and  how this ua.me,Jriend, applied at first only to     misled, who carries one individual in his mind ae the  eign of all he has seen, and all he calculates on seeing,  and reasonB on this one, with a conviction that the  reasoning includes all the others. The idea of an in-  dividual thing which is thus set up as the represent-  ative of a class, may perhaps, without impropriety, be  called a general idea ; and if Locke had never used  the expression but in subservience to such an cxplana-  uon, little or no exception could have been taken to  it. There is a passage (Essay on the Understanding,  Book III., Chap. 3. Sect. Jl.) which perfectly ac-  cords with the doctrine in the text, and proves that  though Locke had misled himself by setting out with  an opinion that the operations of the human under-  standing could be treated of independently of words,  he had more correct thoughts on the subject as he  proceeded. Another passage, giving a correct account  of abstraction with reference to language as the instru-  ment, will be found Book IL Chap. II- Sect. 9-     SKCT. 7.]     109     one who stood in this ration to the speaker,  came at last, by observing and comparing  other cases, to be applied to all who stood in  the same relation to any other person. We  should, in short, have to shew the same pro-  cess with regard to all the examples of modes,  substances, and relations, which Locke's Es-  say supplies; but with these brief hints to  guide him, the reader may be left, in other  instances, to trace the process for himsdf.  It will now be time, — still witii reference to  the principles ascertained in the last chapter,  —to examine some other points of doctrine in-  sisted upon by writers on Logic.   7. The operations of the mind necessary  in Logic are said to be three, viz. Percep-  tion or Simple Apprehension ; Judgment ;  and Reasoning. Under the first of these di-  visions, writers on Logic treat of ideas, or  the notions denoted by separate words, that  is, words not joined into sentences ; — under  the second, they give us separate sentences,  technically called propositions j — ^and under     110     ON LOGIC.     [chap. ir.     the third, they shew how two propositions  may of necessity produce another, so that the  three shall express one act of reasoning. Now,  that perception, judgment, and reasoning,  are all essential to Logic, needs not be called  in question ; but if the theory we have before  us in this treatise be true, the common doc-  trine will appear, by the manner in which it ex-  emplifies these acts of the mind, to have com-  pletely confounded what really takes place, in  the preparation for, and in the exercise of this  art. What, in the first place, is perception but a  sensation or sensations from exterior objects  accompanied by a judgment ? Our earliest  sensations are unaccompanied by any judg-  ment upon them ; for we must have ma-  terials to compare in order to judge ; and  these materials, in the earliest period of our  existence, are yet to be collected. At length,  we can compare j and because we can com-  pare, we judge, and hence we come to know :  — " I know that the object which now affects  my sense of vision is a being like myself; I     SECT. ?•] ON LOGIC. Ill   know him to be one of a great many similar  beings j I know him to be older or younger,  &c. ; I know that what now affects my sense of =  hearing, is the cry or bark of a dog" •, &c.j  I could not know all this, if I had had no  means of judging ; and I can have no means  of judging which the senses do not originally  furnish or give rise to. Perceptiouj then,  (which in every case is more than mere sen-  sation,) always includes an act of judgment ;  and to treat of Perception and Judgment  under different divisions of Logic, must pre-  vent the proper understanding of both. In-  stead, however, of the term Perception, some  writers t use that of Simple Apprehension.  *' Simple apprehension," says Dr. "Wliately,  *' is the notion (or conception) of any object  in the mind, analogous to the perception of  the senses." t The examples appended to     • See Chap. I. Sect. 16.  ■f- Viz. Professor Duncan and Dr. Whately.  J Elements of Logic by Dr. Whately, Chap. II.  Part I. Sect. 1.     [|CHAP. II.     this definition, are, *'inan;" "horse;"  •'cards ;" " a man on horseback ;" " a pack  of cards." Now, if the notion or conception  of tliese, 13 analogous to the perception of  them by the senses, — then, as the perception  includes an act of judgment, so Ukewise  does the conception. But, in truth, the no-  tion corresponding to any of these expressions,  is very different from the perception of a  man, a horse, a man on horseback, &c. ;  and the word or phrase in a detached state  does not stand for a perception or concep-  tion inclusive only of an act of judgment,  but signifies an inference obtained by the use  of a medium, — in other words, a rational  conclusion. For in all cases, what gives the  name and character of rational to a proceed-  ing, is the use of means to gain the end in  view. When we perceive intuitively of two  men, that one is taller than the other, al-  though the judgment we form may be an  e0ect of reason, yet we do not describe it as  a rational process ; but if the investigator,     SECT. 7-] ON LOGIC. 113   not being able to make a direct comparison  between them, introduces a medium, and by  its means infers that one is taller than the  other, then we say the conclusion has been  obtained by a process of reason *. So, in  applying a common name to two individuals  that are intuitively perceived to resemble,  we may be said to exert the judgment, and  nothing more ; but if we apply it to a third,  and a fourth, and a fifth, it is a proof that we  measure each by the common qualities ob-  served in the first two, and that we carry in  the mind a sign of those common qualities  (whether the name, or one of the former in-  dividuals) for the purpose of carrying on the  process. In this way, an abstract word or  phrase, let it signify what it will, provided it  be but abstract, is both the sign of some ra-     • Reasnn is the capacity for using mpdia of any  kind, and it consequent capacity for language : — the  term reasoning has reference to tlie act of thinking,  with the aid of media in order to reach a couclu-     114     [chap. II.     tional conclusion the mind has already come  to, and the means of reaching other conclu-  sions : which statement is true even of a  proper name. For the name John, for in-  stance, underetood abstractedly, does not sig-  nify John as we now perceive him, or as we  have perceived him at any one time ; but it  signifies our knowledge of him separately  from any of those perceptions. But we could  not know of him separately from our percep-  tions, unless we had the power of setting up  some sign (whether the name or aught else)  of what was common to all those perceptions,  and comparing them all with that sign *.   • It is not meant that we could not know him  every time we perceived him, but that we could not  know of him separately from our perceptiong, if we bad  not the power spoken of in the text. It might be  curious to trace this distinction in the case of a dog.  A dog knowE his master every time he perceives him :  — when he does not perceive him, he is reminded of  his absence by some change in his sensations, — (smcU,  for instance, as well as sight, and perhaps some  others ;) he therefore seeks him, and irets if he cannot  find him. But abstracted from all perception, and     SECT. 8.J ON LOGIC. 115   8. It appears, then, from what precedes,  that words and phrases which writers on  Logic give as examples of Perception or  Simple Apprehension distinct from Judg-  ment and from Reasoning, are no examples  at all of the first distinct i'rom the latter two ;  and equally groundless will appear that dis-  tinction which refers a proposition to an act  of judgment separate from reasoning. Not  that an act of reasoning takes place whenever  a proposition or sentence is uttered. For, as  we have seen in the previous chapter, (Sect.  19.) a speaker does not always think of the  separate meaning of the words when he utters  a sentence ; and if a sentence denotes, as a  whole, some sensation or emotion not de-  pendent on reason, (for instance, " My head  aches;" •' My eyes are delighted,") the ut-  tering of it as a whole, without attending to  the sqiarate words, will no moj'e express aa     from all notice by change of sensation, it will scarcely  be contended that a dog knows of his master, as a ra-  tionsl being knows of his absent friend.     116     ON LOGIC.     [chap. II.     act of reasoning, or even of judgment, than  would a natural ejaculation arising out of the  occasion, and used in place of the sentence.  But the following propositions, " Plato was a  philosopher;" "No man is innocent ;" which  are given in Watts's Logic as examples of the  act of the mind called Judgment, stand on a  different footing ; and we affirm that, being  used Logically, they involve not an act of  judgment merely, but express a conclusion  drawn from acts of reasoning.   9- Previously to shewing what has just  been asserted, let us distinguish a grammati-  cal, and an historical understanding of these  sentences ; for a mere grammatical under-  standing of them must be, and an historical  may be, essentially different from the logical  understanding of them. A grammatical un-  derstanding, for example, of the sentence,  Plato was a philosopher, is merely a recog-  nition of its correctness as a form of speech  without considering whether it conveys any  meaning or not ; and it would be grammati-     SECT. 9.3     117     cally understood if any words whatever were  substituted for those that compose the sen-  tence, provided they had a proper syntactical  agreement. An historical understanding im-  plies some concern with the meaning of the  sentence ; but this may be very different in  kind and degree, as depending on the know-  ledge whicli the mind is previously possessed  of. If the hearer did not know what Plato waa  previously to the communication, but knew the  meaning of the word philosopher, he would,  by the sentence, be informed what he was, If  he previously knew, from history, how Plato  lived, thought, and acted, but did not know  the meaning of the term philosopher, the ad-  ditional information conveyed to him by the  sentence, would be but little : he would be in-  formed. Indeed, that he was called a philoso-  pher, but why or wherefore, he could, for the  present, only guess. Let us suppose, however,  that before he comes to calculate why Plato is  called a philosopher, he had heard the word  plied to others : if he bad heard Socrates     m     [chap. II.     called a philosopher, and Confucius a philo-  sopher, he would, on hearing Plato so called,  compafe the individuals in order to ascertain  some common qualities in all, of which the  word might be the sign, and getting these,  he would know or have a notion of the word  philosopher ; though the notion would pro-  bably undergo many modifications as otlier  individuals, Solomon, Seneca, Locke, Rous-  seau, Newton, were successively subjected to  the common sign : — for if the hearer fixes his  notion at once, many individuals will perhaps  be excluded from his class of philosophers,  which other people include under that term ;  and perhaps he will include many, which the  usage of the term excludes. In this way,  then, while our knowledge of what is included  in separate words or phrases is imperfect, we  may nevertheless have some understanding of  the sentences we hear or read ; and this his-  torical understanding suggests the reasoning  process just described, by which we get a  logical understanding of the separate words.     SECT. 10.]     119     10. But now to make a logical use of  tfaem in framing a proposition. We suppose  the preliminary steps, namely the knowledge  included in the separate words ; we suppose  it to be known, from history, how Plato lived,  thought, and acted ; we suppose it to be  known what is meant by philosopfier, by  having heard the word applied to many indi-  viduals i but we have not yet applied it to '  Plato ; in other words, we have yet to ascer-  tain whether Plato belongs to the class of in-  dividuals denominated philosophers. Writers  on Logic talk of a comparison of ideas for  this purpose, and of an intuition or judgment ;  but this, to say the best of it, is an imperfect  and bungled account of the matter. If, in-  deed, to know how Plato lived and acted can  be called an idea, it is necessary to have this  idea ; it is further necessary to have a clear  notion of the term philosopher, — if this again  can be called an idea: — and it is true enough  that in comparing Plato with this sign, we  judge or know their agreement intuitively.     am     ON LOGIC.     [chap.     But out of this intuitive judgment an infer-  ence arises, and the sentence expresses that  inference : a comparison has been instituted  through the intervention of a medium, in  order to ascertain whether Plato is to be as-  signed to a certain class of individuals ; we  intuitively perceive his agreement with the  medium, and draw or pronounce our infer-  ence accordingly, — " Plato was a philoso-  pher." Nor is this the splitting of a hair,  but a real distinction, marked and determined  by that difference in the words so often  pointed out, when understood detachedly,  and when understood as a sentence. The  proposition, Plalu was a pJiilosopher, may be  understood as a whole, without making the  comparison in the mind between what Plato,  and what philosopher, abstractedly signify j  but this, with a full understanding of the  whole sentence, can be done only after the  comparison has once at least been effectually  made : — then indeed, when the comparison  has been made, and the inference drawn, the     8ECT. 11.]     121     sentence which expresses that inference, be-  comes, like any single word, the sign of  ■knowledge deposited in the mind, and, like  such single term, it is fitted to be an instru-  ment of new comparisons, and further con-  clusions.   11. Let us now take another proposition :  *' A philosopher, or every philosopher," (for  the meaning is the same,) " is deserving of  respect." This, hke the other, is an infer-  ence from a comparison which took place in  the mind ; previously to which comparison,  the notion or knowledge included in the word I  philosopher was obtained in the manner lately  described (Sect. 9.) : and the notion included  in the phrase to be deserving of respect was  similarly obtained, but independently of the  knowledge denoted by the other expression ;  — that is to say, the phrase deserving of re-  spect, was originally, we suppose, a sentence  applied to some one thing deserving of re-  spect J whence it was successively applied to  other things till a class was formed — in other     B ON LOGIC. [chap. U.   words, till a notion (knowledge) was esta-  blished in the mind of what things are de-  serving of respect. Now, the present ques-  tion is, whether a philosopher is deserving of  respect ? To determine this, we consider  what a philosopher is, (it is presupposed tliat  we have this knowledge,) and we then niea-  Bure our notion of a philosopher with our no-  tion of what is deserving of respect, and thus  £nd that a philosopher is to be admitted  among the things to which we had been ac-  customed to apply the designation deserving  qf respect : that is to say, we come to the  conclusion, that a philosopher is deserving of  respect. Here, therefore, as before, there has  been a reasoning process previously to the  proposition, and the proposition expresses the  inference from it. And the comparison  having once been made in this instance as in  the other, the sentence becomes, like any  single term, the sign of knowledge deposited  in the mind, and like such single term, is  fitted to be an instrument of new compsrisons,     SECT. 11.] ON LOGIC.   and further conclusions. Well then, we know  from reasoning these two things, that " Plato  IB a philosopher," and that " a philosopher is  deserving of respect." These are detached  WORDS* or sentences : but the mind, in com-  paring them, at once comes to the inference  that Plato is deserving of respect: and the  whole may be expressed in one sentence ;  thus ; " Plato, who is a philosopher, is deserv-  ing of respect j" where Plato-who-is-a-pJiiio-  sopher, is equivalent to a noun-substantive in  the construction of the whole sentence ; and,  deserving-qf-respect is equivalent to another ;  and thus the two, with the assistance of the  verb which signifies them to be a sentence,  are but one proposition. Here, as in the  former cases, a comparison has been made \ij.  means of the signs of deposited knowledge ^  for we knew that Plato was a phUosopher;  we knew a class of things or persons deserv-  ing of respect: — comparing our knowledge by   • See the second note (Aristotle's definition of a'  vord bcuig the first) appmded to Sect. 20. Chap. I.     324 ON LOGIC. [chap. ir.   means of the sign deserving-of-respect, the in-  ference follows, that " Plato, who is a philo-  sopher, is deserving of respect." And the  comparison having once been made in this  instance as in the others, the sentence be-  comes, like any single terra , the sign of know-  ledge deposited in the mind, and either in  this or any other equivalent form, is fitted to  be an instrument of new comparisons and  further conclusions. And in this manner are  we able, ad infinitum, to investigate new  truths by means of those already ascertained,  always making use of former words or their  equivalents, as the means of operation.   12. Now, so far as Logic is the art of in-  vestigating truth, (and we intend to show that  its office ought not to be considered of further  extent,) this is the whole of its theory. We  have defined it as the right use of words with  a view to the investigation of truth ; and the  way in which words are used for the purpose,  is that which has been described : — in brief,  they are used by the mind in making such     SECT. 19,] o^f LOGIC. 125   comparisons as it cannot make intuitively. Of  two objects, or of a sensation or emotion  twcie experienced, we can intuitively judge  what there is in common between them;,  l< suppose a third object, or a sensation, &c«  thrice experienced, an intuitive judgment can  still be applied only to two at a time, and wei  can but know in this way what there is  common to every two. But if we set up tf  sign of what is common to two, we can compare  with the sign a third, and a fourth, and a  fifth, and judging intuitively how far it agrees  with the sign, we infer its agreement in thq  same proportion with the things signified,  In Logic, the sign used is always presumed  to be a word. Now, in our theory of Ian-  guage, every word was once a sentence ; and  every sentence which does not express the  full communication intended, but is qualified  by another sentence, or becomes a clause of a  larger sentence, is precisely of the nature of  any single word making part of a sentence *.  • See Chap. I. Sect. 28.     IM     I^CMAP. 11,     From the first moment, then, of converting  the expression used for a particular communi.  cation, into an abstract sign of the sentiment  or truth which that communication conveyed,  the mind came into possession of the instru-  mental means for furthering its knowledge :  and this means always remains the same in  kind, and is always used in the same way.  The word which once signified a present par-  ticular perception, ceased, through the ne-  cessities of language, to signify that percep-  tion in particular, and came to signify, in the  abstract, any perception of the same kind, or  the object of any such perception. In this  state, it no longer communicated what the  mind felt, thought, or discovered at the  moment, but was a sign of knowledge gather-  ed by comparisons on the past. By u«ng this  Bign, the mind was able to pursue its inves>  tigations, and every new discovery was de-  noted by a sentence which the sign helped to  form, its general application being limited to  the particular purpose by other signs. But if     SECT. 13.]     ON LOGIC.     137     one WORD"     ' may lose its particular pnrpose,  and become an abstract sign, so may another,  and be the means, in its turn, of prosecuting  further truths, and entering into the com-  position of new WORDS. Thus will the procesa  which constitutes Logic, be aiways found one  and the same in kind, having for its basis the  constitution of artificial language, such as it  was ascertained to be in the previous chapter.   H 13. Now of this Lc^ic, — the Logic, uni-   H versally, of ntpotres, or woKD-dividing men, —   H let the characteristics be well observed, in order   H to keep it clear from any other mode of using   H signs for the purpose of reasoning, to which   H the name of Logic is attributed. The Logic   H here described, is a use of words to regista-   H our knowledge as fast as we can add to it, by   H new examinations, and new comparisons of   I things } each new esamination, each new   H sen!     • The reader will bear in mind the comprehenBive  sense of the term which we have in view, when it is  printed in capitate.     US*' ON LOGIC. [chap. II.   comparison, being made with the help and  the advantage of our previous knowledge.  The reasoning takes place in the mind in such  a manner that it is not a comparison of terms,  but a comparison of what we newly observe,  with what we previously knew. Words indeed  are used, because without signs of one kind  or of another to keep before the mind the  knowledge already gained, we could compare  only individuals j but however words may in-  tervene, it is always understood that the mind,  at bottom, compares the things, A man  may be informed, that, " Plato who is a phi-  losopher, is deserving of respect;" that,  " William who is recommended to his service,  is an honest man ;" that, *• A particular tree  in his garden, is a mulberry tree ;" that,  " Stealing is a vice, and temperance is a  virtue ;" that, " Throughout the Universe, all  greater bodies attract the smaller ;" that, " A  triangle described within two circles in such  a manner that one of its sides is a radius of  both, and the others, radii of each circle     SECT. 13.]     129     respectively, is an equilateral triangle;" — a  man may be informed of these and similar  ^'things, and may entirely believe the inform-  ation; nay, hemayjustifiably believe it J for he  may know of those who give it, that their ho-  nesty is such, that they would not wilfully de-  ceive him ; that their intelligence and inform-  ation are such, that they are not likely to say  what they do not know to be true : but a man  can be said to know these things of his own  knowledge, and in this way to be convinced  of their truth, only by a process of reasoning  that musl take place within his own mind ; a  process which can take place only in a mind  by nature competent to it, and which requires,  in every case, its proper data or facts, aided,  it is true, by language, or by signs such as Ian-  guage consists of, to register each inference *,  • The necessity of language, as a means of in-  vestigation, applies not to our last example. The mincl  may investigate (though no one can demonstrate)  mathematical truths, with no other aid than visible  diagrams ; or even diagrams that are seen only by  " the mind's eye."     130 ON LOGIC. [chap. II.   and so to get from one inference to another,  and thus, ad infinitum^ toward truth. Be-  cause the several steps, leach of which is a  conclusion so far attained, cannot take place,  without the instrumentality of signs to assist  the mind, we consider the process an art ; and  if the signs used are words, the art is pro-  perly called Logic. But whatever aid the  reasoner may borrow from words, the only  true grounds of his knowledge are the facts  about which the reasoning is employed.  Without them, no comparison of the terms  can force any conviction further than that  the terms agree or disagree. He may be told  that — " Every philosopher is deserving of  respect,*' and that, — " Plato is a philosopher :**  but if he knows not what a philosopher is, or  what it is to be deserving of respect, the  comparison of the terms in order to draw a  conclusion from them, will be a mockery of  reason : — it will be reasoning indeed, but  reasoning without a rational end. And suppose  the knowledge to have been acquired of what  a philosopher is by the application of the word     SECT. 13.] ON LOGia 131   to many particulars, and by a consequent  classification of them in the mind, — supposing  the knowledge of what is deserving of respect  to have been acquired in the same way, —  supposing the inquirer has learned from history  what Plato was in his opinions and manner of  life, — the conclusion takes place by a com-  parison of the thingSj by means indeed of  words, but not by any comparison of the terms  independently of the things ; nor is the con-  viction in the least fortified, or the process ex-  plained, bya demonstration that in reasoning  with the terms alone, independently of their  meaning, we get at the conclusion ; — by  shewing, for instance, that the terms which  include the facts, may be forced into cor-  respondence with the following ^nwwfa;  Every B is A :  C is B :  Therefore C is A.  Every philosopher — is— deserving of respect :   Plato — is— a philosopher :  Therefore Plato — ^is — deserving of respect.   K 2     .18«     ON LOGIC.     [chap.     This way of drawing a conclusion from a  comparison of terms, is. properly speaking, to  reason or argue with words ; but in the Lo-  gic we have ascertained, every conclusion is  required to be drawn from a comparison of  the facts which the case furnishes ; and words  being used only for the purpose of registering  our conclusions, such Logic is properly de-  fined the art of reasoning by means of words.  The inquirer who seeks to know, of his own  knowledge—" Whether William who is re-  commended to his service, is an honest man",  — will gather facts of William's conduct by  his own observation ; and these he will com-  pare by the light of his previous notion (i. e.  knowledge) of what an honest man is : but  then he must have that previous notion, or he  cannot make the comparison ; and the notion  will have been gained by a process just like  that he is pursuing : and so downwards to the  original comparison of individiial tJujigs, from  which all knowledge begins. So again, if an  inquirer seeks to know that " a particular tree     SECT. 13.] ON LOGIC. 133   is a mulberry tree", — he must first know  what a mulberry tree is; and how can he  know this but by a comparison of different  trees? There must be some art employed to  classify the individual trees, otherwisehe could  never know more than the difference between  every two trees. By setting up one tree, or  some equivalent sign, as a word, to denote  the common qualities observed in many, he  comes to know what a mulberry tree is ; and  looking at the particular tree in question, he  sees that it has the common qualities indica-  ted by the sign, and infers that it is a mul-  berry tree. So likewise, if an inquirer seeks  to be convinced that " SteaUng is a vice",  or that "Temperance is a virtue", — he  must have such facts before him as will  enable him to come to a clear conclusion as  to what is vice, and what is virtue : and  this conclusion will either include or ex-  clude stealing with respect to his notion  of vice, and temperance with respect to his  notion of virtue, and he will consequently be     134     [chap. [I.     convinceti or not convinced of tlie proposition  in question. So, once more, if an inquirer  desires to know, of his own knowledge,  *' Whether, throughout the universe, all  greater bodies attract the smaller", — he must  first observe certain facts from which the ge-  neral law may be assumed hypothetical ly : —  he must then ascertain what, according to  other notions gained from experience, would  be the effect throughout the universe of the  general law which he has so assumed ; and if  the effects arising out of the hypothesis cor-  respond with actual effects, and no other by-  pothesis to account for them can be framed,  he will have all the proof the subject permits,  and know of his own knowledge, as far as can  be known, the conclusion asserted. So, lastly,  if an inquirer seeks to be convinced that "a  triangle described within two circles in such  a manner that one of its sides is a radius of  both, and the others radii of each circle re-  spectively, is an equilateral triangle", — he  must first form within his mind the notions of     SECT. 13.] ON LOGICi tS5^   a triangle, and of a circle, the latter of which he  will find can be conceived perfect in no other  way than in correspondence with this definition :  — "a plane figure bounded by one line called-  the circumference ; and is such that all straight  lines, (called radii,) drawn from a certain  point within it to the circumference, are equal  to one another. " Having formed this notionr^  he will find, by certain acts of comparison^  (which must take place within the mind, al-  though they may be attsisted by a* visible sign-J^  that the previous proposition is an inevitable  consequence of the notfon so formed, and his'  conviction: wiU be comffiete. If the convic-  tion, in the previous ifrstances, has not the  same force as iiti the last^ — ^if, in those instances,  the force may be diffident m. degree, while in  the last there can be no coD^victioa short of  lliat which iS' absolute an4- entire, the cause^  in not that the reasoning process^ is different  in kind, but that the facts or data about which"  it is' employed are dii&re»t. In the last in^  stance^ the reasoning is employed about no-     136     ON LOGIC.     [chap. II.     tions, which admit uf being so defined, that  every mind capable of the reasoning at  once assumes them before the reasoning pro-  cess begins ; but in the other instances, the  facts or the notions may be attended by cause  for doubt. A man, if he have any notion of  a philosopher at all, cannot indeed but be  quite sure (consciously sure) of his own no-  tion of a philosopher j but how can he be sure  that others have the same notion, or even  quite sure that Plato had the qualities that  conform to his own notion ? In the same  way, he will be quite sure (consciously sure)  of his own notion of an honest man ; but he  may be deceived as to the facts which bring  William within that notion. He will be quite  sure (consciously sure) of the notion he has  in naming a tree a mulberry tree ; but that  notion may be totally unlike the notion which  other people entertain ; or if the general no-  tion agrees, he may mistake the characteristics  in the particular instance. He will be quite  sure (consciously sure) of his own notion of     SECT. 13.]     137     vice or of virtue, and whether it includes or  excludes this or that conduct, action, habit,  or quahtjr ; and in this case the conviction is  absolute and entire while the reasoner confines  himself to his own notion ; but the moment  he steps out of this, and begins to inquire  whether it agrees with that of others, he finds  cause to doubt. He must be quite sure (sen-  sibly sure) that bodies near above the earth's  surface have a tendency towards it ; and by  proper experiments he may convince himself  that all bodies without exception which are  so situated, have the same tendency. In sup- ,  posing the fact universal of the tendency of  smaller bodies to the greater, his conviction  of the consequences involved in that hypo-  thesis, must, as soon as he has mentally traced  them, be absolute and entire ; but he has yet to  find whether reality corresponds with the hy-  pothesis. The strongest proof of this will  be, the correspondence of the consequences of  the hypothesis with the phenomena of na-  ture, joined to the impossibility of forming     138 ON LOGIC. [chap. II.   another hypothesis which shall account for  these phenomena; and the doubt, if any,  will attach to that impossibility, and to the  accuracy of bis observatioda of the pheno*  rneoa* I^ then, there is roonr for doubt, and  cocise^aently for various degrees of assent, in  all the instances except m that whose facts or  data are notions which the mind is bound to  tstke up according to the definitions before it  enters on the argument, we are not to con-  clude that the reasoning process is different in  kind iti any of them ; since the difl^ence in  the facts or data about which the reasoning  process i& employed, fully accounts for the ab-  solute and entire conviction which takes place  in one instance, and the degrees of convictioti  which are liable to happen in such cases as^  the others.   14. But what IB a process or act of rea^  soning? Is it, abstractedly from the means'  u£^d to register its conclusions, and so pro-  ceed to new acts of the same kind, — ^is it aa  act which rules can teach, or any generalbsau-     SECT. I4.j     139     tion make clearer, or more satisfactory than it  is originally ? We shall find, upon examina-  tioH, that any such pretence resolves itself in- i  to a mere verbal generalization, or the appli-  cation of the same act to itself; and that this  does in no way assist the act of reasoning, or  explain, or account for, or confirm it. A man  requires not to be told — *' It is impossible for  the same thing to be and not to be," in order  to know that himself exists ; he requires not  the previous axiom, " The whole is greater  than its part, or contains its part, " in order to  know that, reckoning his nose a part of his  head, his head is greater than his nose, or his  nose belongs to his head ; neither is the previous  axiom, " Things equal to the same, are equal  to one another", necessary to be enounced,  before he can understand, that if he is as tall  as his father, and his father as his friend, he  is as tall as his friend *. Whatever neatness of  arrangement a system may derive from being   • Compare Lofku's Essay, Book IV. ChajHeis 7  and 12.     140     ON LOGIC.     [chap. II.     headed with such verbal generalizations, it is  manifest that they neither assist the reasoning  nor explain it : nor must a generalization of   , this kind be confounded with the enunciation  of what is called a law of nature*, — (the law  of attraction and gravitation for instance, — )  since this last is a discovery by a process of  experiment and reasoning, but a verbal gene-  ralization is no discovery at all ; — it is merely  a mode of expressing what is known by every   " rational mind at the very first opportunity for  exercising its powers. Or more properly  speaking, the laws of reasoning, which are  gratuitously expressed by what are called  axioms, are nothing else than a mode of de-   * See Whately's Logic, Chap. I. Sect. 4, where he  attempts to evade Dugald Stewart's oh^ection to the  Ariatotelian syllogism, that it is a demonstration of b  demoiigtration, by comparing the Dictum de omni et  de nullo to the enimciation of a law of nature. — It is  rather pleasant, in the first note of the Chapter referred  to, to hear the doctor running riot upon Locke's con-  fuinon of thought and common place declamation, be-  cause the latter had the sense to sec the futility and  puerility of the syllogism.     SECT. 14.] ON LOGIC. 141   scribing the constitution of a rational mind.;—*  they are identical with the capacity itself for  reasoning: to view them in any other light is  to mistake a circumlocution for the discovery  of a principle. And this kind of mistake  every one labours under who supposes that,  by any means whatever, an act of reasoning  is assisted or explained, accounted for, or con-  firmed. Nothing is more certain, than that if  two terijns agree with a third, they agree with  each other, — if one agrees and the other dis-  agrees, they disagree with each other: but  every other act of reasoning has a conclusion  equally certain (the facts or data about which  an act of reasoning is conversant being the  sole cause of any doubt in the conclusion*,)  and this or any other attempt at explaining or  accounting for the act, will therefore only   . * And note, that when people are said to draw a  wrong conclusion from facts, the correct account would  be, that they do not reason from them, but from some-  thing which they mistake for them, through their ina-  ability to understand, or their carelessness to the na-  ture of, the facts given.     I4!l     [chap. ir.     amount to the placing of one such act by the  side of another; as if any one should set a  pair of legs in motion by the side of another  pair, and call it an explanation of the act of  walking. Such would at once appear to be  the character of the Aristotelian Syllogism,  were it not for the complicated apparatus ac-  companying it ; an apparatus of distinctions  and rules rendered necessary by the nature of  the terms compared. For these terms being  obtained by the division of a sentence, are  such that they agree or disagree with each  other only in the sense they bore before the  division took place. Our theory makes this  plain; for it shows that words which form a  sentence limit and determine each other, and  thus have a different meaning from tliat which  belongs to them when understood abstracted-  ly. Therefore, though it may be true that  " Plato is a man deserving of respect, '  does not follow that " Plato " and " A maai  deserving of respect " shall agree togetiier as  abstract terms : accordingly the latter term     SECT. I'i.]     143     understood abstractedly, signifies any or every  man desei-ving of respect, and does not agree  with Plato. It must be obvious, then, that  terms obtained iirthis way, can be compared  with other terms similarly obtained, only un-  der the safeguard of certain rules. Such rules  are accordingly provided ; and tliat they may  not want the appearance of scientific general-  ization and simplicity, they are all referred to  one common principle, — the celebrated dic-  tum de omni et de nullo ; whose purport is,  that what is affirmed or denied of the whole  genus, may be affirmed or denied of every  species or individual under it ; — which indeed  is nothing more than a verbal generalization  of such a fact as this, that what is true of every  philosopher, is true of any one philosopher.  All tliese pretences to the discovery of a uni-  versal principle, do but leave us just where we  were, a few high-sounding empty words ex-  cepted; and this must ever be the case when  we seek to account for that which is, by the  constitution of things as far aa we can ascer-     ON LOCTC.     [CHA     tain them, an ultimalefact. An act of reason-  ing is the natural working of a rational mind  upon the objects, whatever they may be, which  are placed before it, when, having formed one  judgment intuitively, it makes use of the re-  sult as the medium for reaching another: and  the pretence to assist or explain this operation  by the introduction of such an instrument as  the syllogism, is an imposition on the under-  standing.   15. This will more plainly appear when we  examine the real use, (if use it can be called,)  of the Aristotelian art of reasoning. It may  be described as the art of arguing unreason-  ably, or of gaining a victory in argument  without convincing the understanding. As  it reasons "with words, and not merely by  means of words, it fixes on expressions not on  things, and is satisfied with proving a conse-  quence, or exposing a non-sequitur in those,  without inquiring into the actual notions of  the speaker. " Do you admit " says a syllogi-  zer, " that every philosopher is deserving of     SECT. 15.] ON LOGIC. 14.5   respect? " " I do;" says the non-syllogi-  zing respondent. " And you admit, (for I  have heard you call him by the name,) that  Voltaire is a philosopher : you admit, there-  fore, that Voltaire is deserving of respect. "  Now, if the notion of the respondent is, that  Voltaire is not deserving of respect, here is a  victory gained over him in spite of his con-  viction. Arguing from the words, and allow-  ing no appeal from them when once conceded,  the conclusion is decisive*. But in looking  beyond the words to the things intended, we  shall find that the respondent either did not  mean every philosoplier, as a metaphysical,  but only as a moral universal, or else (and the  supposition is the more likely of the two) that  in calling Voltaire a philosopher, he called   • " If," says a. doughty Aristotelian doctor, " a  imiyeraity is charged with cultivating only the mere  elements of mathematics, and in reply a list of the  hooks studied there is produced, ^should even any one  of those books be not elementary," [" / day here on  my biynd,''] " the charge is in fiiirncss refuted."  Whately's Logic, Chap III. Sect. 18.     146 ON LOGIC. [chap. II.   him so according to the custom of others, and  not according to his own notion. In a Logic  whose object is truth and not victory, the  business would not therefore end here. An  attempt would be made to change the notion  of the respondent (supposing it to be wrong)  by an appeal to things. His mind might in-  deed be so choked with prejudice as to be in-  capable of the truth ; but at least would the  only way have been taken to remove the one  and procure admission for the other. — To the  foregoing, let another kind of example be add-  ed : " Every rational agent is accountable ;  brutes are not rational agents ; therefore, they  are not accountable." * " Non sequitur*^  cries the Aristotelian respondent. The other  man, who reasons by means of words and not  merely mth words, is certain that the internal  process by which he reached the conclusion is  correct ; nor is he persuaded to the contrary,  or at all enlightened as to his fault, when he  is told that he has been guilty of an illicit pro-   ♦ From Whately's Logic, Chap. I. Sect. 3.     SECT. 15.] ON LOGIC. 147   cess of the major. He is informed, however,  that his mode of reasoning finds a parallel in  the following example : " Every horse is an  animal ; sheep are not horses ; therefore they  are not animals.'* * But this he denies ; be-  <:ause he is sure that his mode of reasoning  would never bring him to such a conclusion  as the last. All this time, while the Aristo-  telian has the triumph of having at least  puzzled his uninitiated opponent, the real  cause of diflference is kept out of sight, name-  ly, that the one refers to that reasoning which  is conducted merely with words, and not by  means of words only, while the other refers to  that reasoning which looks to things, inatten-  tive perhaps, as in this instance, to the expres-  sions. If the latter had used no other ex-  pression than " Brutes are not rational agents ;  therefore they are not accountable ;•" — the as-  sertion and the reason for it, must have been  suffered to pass; but because another sen-  tence is prefixed to these two, and the whole   * Whately'*s Logic, Chap. I. Sect. 3.   l2     F   1     148 ON LOGIC. [^CHAP. II.   of them happen to make a violated syllogism,  the speaker is charged with having been guilty  of that violation, when in fact he has not at-  tempted to reason syllogistically at all ; i. e. to  draw his conclusion from a comparison of the  extremes with the middle, but from a judg-  ment on the facts of the case. In a Logic  which gets at its conclusions by jneans of  words, and not by the artifice we have just  referred to, an expression which does not  reach the full facts reasoned from, (every  rational agent, for instance, where it should  have been said none but a rational agent,J  would not be deemed an error of the rea-  soning, but a defect in the expression of the  reasoning.   ] 6. These examples will, it is hoped, be  sufficient to show the real worth of the Aris-  totelian syllogism, ft is indeed, as its advo-  cates assert, an admirable instrument of ar-  gumentation ; but of argumentation distinct  from the fair exercise of reason. It is a pro-  per appendage to the doctrine of ReaUsm,      SECT. 16.]]     149     and with that exploded doctrine it should long  ago have been suffered to sink. While ge-  nera and species were deemed real independ-  ent essences, to argue from words was con-  sistently supposed to be arguing from things :  but now that words are allowed to be only  counters in the hands of wise men, the Logic  of Aristotle, which takes them for money,  should surely be esteemed the Logic of fools".  The claim for its conclusions of demonstrative  certainty, rests solely on the condition that  words are so taken. Every conclusion from  an act of reasoning, would have that charac-  ter, if the notions about which it was employ-  ed were notions universally fixed and agreed  upon. In mathematics, this circumstance is  the sole ground of the peculiar certainty at-  tained. All men agree in the metaphysical  notion of a point, of a line, a superficies, a  circle, and so forth t : if all men necessarily     * " Words are the counters of wise men, but the  money of fools," — Hobbes.   f According tu Sugald Stewart, mathematical     IW ON LOGIC. [chap. il.   agreed in the notion of who is a philosopher  and who is not, of what is vice and what is  virtuBj and so forth ; our conclusions on these  and similar subjects, would, as in mathematics,  be demonstrative : but till definitions can be  framed for Ethics in which men must agree,  there is little chance of erecting this branch  of learning, with any praciical benefit, into a  science, according to the notion insisted on  with some earnestness in Locke's Essay*,  lu Physics we can do more ; for men agree  pretty well as to what is a mulberry tree, and  what is a pear tree ; what is a beast, and what  is a bird ;— by experiment they can be shewn  what are the component parts of this sub-  stance, what the qualities of the other j and  so forth : so that here, our conclusions need   definitions are mci-e hypotheses. Do they not rather  describe notions of and relating to quantity, which, by  the congtitution of the mind, it must reach, if, setting  aside the sensible instances of a point, a line, a circle,  &c., it tries to conceive them perfect ?   * Book IV. Chap. III. Sect. 18,: and the same  book Chap. XII. Sect. 8.     SECT. 17.] ON LOGIC. 151   not be wanting in all necessary certainty;  although, as that certainty depends on the  conformity between our notions, and the out*  ward or sensible objects of them, it will be of  a different kind from the certainty obtained  in meta-Phi/sicSj and therefore not called de-  monstrative. In the latter department, (Me-  taphysics,) the chain of evidence has its first  hold, as well as every subsequent link, in the  mind, and the mind cannot therefore but be  sure of the whole.   17. As we propose to limit the province  of Logic to the investigation of truth, the re-  marks and examples in the section preceding  the last (15.), might have been spared till we  come to consider Rhetoric, to which we in-  tend to assign, among its other ofiices, that  of proving truth. How far the form of ex-  pression which corresponds to the syllogism,  is calculated to be useful to a speaker or wri-  ter, may at that time draw forth another ob-  servation on the subject. Meanwhile we pro-  pose to exclude it entirely from Logic; and     U3     ON LOGIC.     [chap, II,     in truth the common practice of manlcind out  of the schools, has never admitted it as an in-  strument either for the one purpose or the  other. Common sense has always been op-  posed to it ; and Logic is a word of bad reputa-  tion, because it is supposed to mean the art  of arguing for the sake of victory, and not for  the sake of truth. In vain have Locke,  Campbell, Reid, Stewart, and other sound  thinkers, endeavoured to clear the art from its  reproach by detaching the cause : the Aristo-  telian Syllogism has been repeatedly over-  thrown ; yet some one is ever at hand to set it  on its three legs again, and argue in defence  of the instrument of arguing : — some per-  tinacious schoolmaster may always be found  Who e'en though vanquished yet will ahgue still;  While words oflearncd length and thundering sound*.  Amaze the gazing rustics ranged around.     * Videlicet, Terms middle and extreme ; premiss  major and minor ,- quantity and quality of propositions ;  Universal affirmative ; Universal negative ; Particular  affirmative ; Particular negative ; Distribution and non-  distribution of terms; Undistributed middle; Illicit pro-     SECT. 18.]     ON LOGIC.     So much — (till, in the next chapter we come  to a parting word — ) so much for the Aris-  totelian Syllogism.   18. As to the Logic which we have en-  deavoured to ascertain, it is, we repeat it, the  Logic which all men learn, and all men ope-  rate with in gathering knowledge ; and the  only inquiries which remain are, i. Whether,  so far as we have gone, there is ground or ne-  cessity for principles and rules in the exercise  of Logic, as there is for grammar in speaking  a language; and ii. Whether we ought to  consider its limits as extending beyond the     cBss of the major ; Illicit piocese of the Tninor ; Mood  itnd figure— Barbsrs, Celarent, Darii, Ferio, Cesare,  CameBtres, Festino, Baroko, Darapti, Disamis, Datisi,  Felapton, Bokardo, Feriso, Bramantip, Camenes, BU  maris, Fesapo, FrcBison ; Categoricals, Modals, Hypo-  theticals. Conditionals, Constructive form. Destructive  form, Oatcnsive reduction, Illatire conversion, &c. kc  &c. Well may we join with Mons. Jourdain —  " Voila dee mots qui sont trop rebarbatifs. Cette  logique ]& ne me rcvient point. Apprcnons autre chose  qui soit plus joli.'*     lAt ON LOGIC. [chap. II.   bounds proposed at tlie commencement ot*  this Chapter.   19. Though few persons would be dis-  posed to answer the former question in the  negative, yet an analogous case may induce a  moment's pause in our reply. At the conclu-  sion of the first note appended to Sect. 4.,  allusion was made to the fact, that men do  not see truly by nature, but acquire, through  judgment and experience, the power of know-  ing by sight the tangible qualities of objects  and their relative distances. Now, the in-  terference of rules, supposing them possible,  to assist this early discipline of the eye, would  be useless — perhaps raiscliievous : — why are  we to think differently of the discipline of the  mind, as regards the use of those signs which,  if our theory is true, are forced upon us at  first by an inevitable necessity ? Because the  art of seeing truly is necessary to the preserva-  tion of the individual ; and nature takes care,  therefore, that we do not teach ourselves im-  pertectly or erroneously ; but the conducting     SECT, ly.] ON X-OGIC. 155   of a train of reasoning with accuracy and pre-  cision into remote consequences, is unne-  cessary in a rude state of society j and man,  who is left to improve his physical and moral  condition, has the instrument of that improve-  ment confided to his own care, that he may  add to its powers, and form for himself rules  for using it with much more precision and  much more effect, than any random use of it  can be attended with. Accordingly, if we  look to that department of knowledge which  Locke calls ipvaiK^ * , we shall find that it owes  its existence to the accurate Logic by which  inquirers registered all their observations and  all their experiments, and by which they as-  cended from individuals to classes, till each  had comprehended in his scheme all he de-  sired to consider. Here then begins the pro-  per business of Logic as a system of instruc-  tion : it ought to lay open all the various me-  thods of arrangement and classification by     ' Vide the lutrixluction to this Treatise.     XISS ON LOGIC. [^CHAP. 11.   which science is acquired and enlarged ; and  if something may yet be done toward im-  proving these methods, it should open the  way to such improvement. The Aristotelian  rules for definition, which are a sound part of  Logic, should be explained and illustrated ;  and the nomenclatures invented by various  philosophers, particularly that which is used  in modern chemistry, should be detailed and  investigated.   SO. But if, by the application of a more  accurate Logic than belongs to a random use  of language, men have been able to accom-  plish so much in ^uo-ik^, it does not appear  that they have great cause to boast of their  success in the other department, namely  ■n-paKTiK-^. Do they act, whether as com-  munities or individuals, muck better with a  view to their real interests, than they did two  thousand years ago ? If improvement here,  as in the other department, is possible, how  is it to be accomplished ? We live in an at-  mosphere of passions, prejudices, opinions,     SECT, go.] ON LOGIC. 157   which mould our thoughts, and give a cer-  tain character and hue to all the objects of  them ; — these we do not examine, but take  them as they appear to us, and our reasonings  too often start from them as from first facts.  As to the process itself, — a process which  every individual conducts ■within his avra  mind according to the power which nature  gives him, — we affirm that it cannot be other  than it is, and that, provided it starts from  true data, it can never lead us wrong : but if  that is false which at the outset we take for  true, then indeed our conclusions may be  perniciously, ruinously erroneous. It is ac-  cordingly the business of the moralist to re-  move the false hue which habit, opinion, and  passion, cast over the surface of things ; and  it should be the business of the politician to  examine the principles on which the general  affairs of the world are conducted, and open  the eyes of mankind to their pernicious ten-  dency, if in the whole or in part they are per-  nicious. But neither the moralist nor the     158 ON LOGIC^ []CHAP. II.   politician can come at the necessary truthis  intvitiveljf : they must use the mediaj and the  media consist in that use of words which con-  stitutes Logic, as we have described it. We  do not intend to say that language affords  the means of reaching equal results to every  person who makes the right logical use of it ;  for men's minds are very different in natural  capacity; and some are able to perceive  truths intuitively, which others attain only by  a slow process; as tall men can reach at  once, what short men must mount a ladder  to : but we do intend to say, that, let the  natural powers of any human mind be what  they will, there is no chance for it of any ex-  tensive knowledge, but through the employ-  ment of media to assist its natural operations ;  <and, we repeat it, the media which nature  suggests, and leaves for our industry to im-  prove, is language *. Well then, if our im-   * The reader does not understand us, if he  deems it an objection to our reasoning, that many  highly gifted men in point of understanding, do not     SECT. 20.] ON LOGIC. 159   provement in ntpaKrucrfj is, at this time of ^ay,  less than we might expect, is it not reason-  able to think that, with regard to this depart-  ment, we do not quite understand the instru-  mental means, and consequently do not ap-  ply them with complete effect ? Surely there  is some ground for such a suspicion, when we  find a doctor (of some repute we presume) in  one of our two great places of learning, de-  claring that '^ the rules of Logic have nothing  to do with the truth or falsity of the premises,  but merely teach us to decide (not whether  the premises are fairly laid down, but)   appear to have a skilful use of language. A man may  be rhetorically unskilful in language without being  logically so ; — he may be imable to convey to others  how and what he thinks ; but he may make use of  media in the most skilful manner to assist his own  thoughts. And if his capacity is such that he seei  many truths intuitively for which others require  media^ it is evident that he cannot convey those  truths to them till he has searched out the means.  The nature and the principle of such an operation be-  longs to our next chapter on Rhetoric.     fim     ON LOGIC.     [chap. 1     whether the conclusion fairly follows from  the premises." * We acknowledge that the  Logic to which this description applies, has  never been the Logic of mankind at large,  however it may have been the baby-game of  men in colleges ; but that the office of Logic  should be described so completely opposite  to what it really is, at a time when its proper  office and character ought to have been long  ago thoroughly understood, is not a little  surprising, and may reasonably warrant the  suspicion stated above. We have no doubt  our reader is by this time convinced, that  men who reason at all, do not want rules for  drawing their conclusions fairly, if we could  but get them to draw those conclusions from  right premises ; and that to get at right pre-  mises is every thing in Logic. For this end,  it is our business to set all notions aside that  have not been cautiously acquired ; and to  begin the formation of new ones at the point   * Whateiy'a Logic. Provinceof Reasoning, Cliap-  I. Sect. 1.     sf;ct. 20.]     IGI     where all genuine knowledge commences, —  the intuitive comparison of particulars or  single facts ; to make use of the knowledge  (notions) hence obtained as media for new  comparisons or judgments; and so on ad in-  Jinitum. Alas! it is but too certain, that  though we draw our conclusions faiily enough,  our premises, in a vast proportion of cases,  are laid down most foully, because they are  laid down by our ignorance, our passions,  and our prejudices ; and because language  itself, when its use is not guarded, is a means  of deception*.   • We arc somewhat backward in offering examples  of general remarks, such as is this last ; because it is  scarcely possible to be particular without touching on  questions in religion or politics that carry with them,  either way, a taint of parti zanshi p ; and we hold it to  be very impertinent in a writer on Logic, to turn  those general precepts for the discovery of truth  which he is bound to ascertain, into a particular chan-  nel in order to serve his own sect or party. What  business had Watts to exempliiy so many of hU  cautionary rules by the errors of Papistical doctrine,  at a time when its doctrine was a subordinate and     '16S ON LOGIC. [^CHAP, II.   21. But can the assistance which lan-  guage is intended to furnish, be rendered such   party queBtioit, and be himself was a sectarian opposed  to it ? We trust that no exception of the same kind  can be taken {particularly as we give them only in a.  note) to two examples we are about to submit of  the remark in the text, that language itself may lie  the means of deceiving us into wrong premiseB : — they  are by no means singular, hut Guch as may he met  with every hour on almost every question. The  ph rase natural state is, as we all know, a very com-  mon expression, which we are much in the habit of  applying to things that have not been abused or per-  verted from the form or condition in which nature  first placed them. Now, because the same phrase  happens to be frequently applied to man in a rude  state of society, we start, in many of our reasonings,  with the notion, that in proportion as we have depart-  ed from such a state, we have perverted and abused  the purposes of nature ; when, in truth, it seems wiser  to inquire, whether we have yet reached the state  which nature means for creatures such as we are, and  whether she is not constantly urging us on to such an  unattained state. Our other example is of narrower in-  terest, and belongs to politics, or rather to what is  called political economy. The word price, in general  loose speaking, means that which is given (be it what  it may) to obtain some other thing ; but in a strict     SECT. 21.] ON LOGIC. 163   as to lead us to truth in spite of ignorance,  passion, and prejudice, and in spite of the  delusions of which it is itself the cause? Why  not, if the guarded and careful use of it, is  fitted to diminish these obstacles, and if we  do not look for the ultimate effects -faster  than, by the use of the means, the obstruc-  tions ^ive way ? Nor are mankind inattentive  to improve the means, nor are the means     and mercantile Bense, it has a uniform reference, direct  or indirect, to the quantity of precious metal given for  commodity ; inasmuch as gold and silver are the sole  universal medium of barter throughout the world, and  every promise to pay has reference to a certain quan-  tity of one or the other of these metals. These things  premised, it must be obvious that the phrase price of  gold, using price in a strict sense, is an abeurdity, and  could arise only from confounding the meaning which  prevails in ordinary speech with the meaning in which  the merchant uses it. What, then, are we to think of  an English House of Commons, which, some twenty  years ago, deputed to a committee the task of in-  quiring into the causes of the high price of bullion ?  Might not the committee, with as much reason, have  been deputed to inquire, why the foot rule was more  or less than a foot ?     164 ON LOGIC. [chap. II.   without effect : for when we ask, whether their  moral and political condition is much ad-  vanced beyond what it was in the most pro-  mising state of the world in past days *, we do  not mean to deny what every one of common  knowledge and observation is aware of, that  it has advanced : all we urge is, that a sys-  tematic attention to the means of investigating  truth, might, peradventure, in politics and  morals, as it has in physics, have been at-  tended with effects more widely beneficial.  Neither do we afSrm that existing works on  Logic are destitute of many admirable pre-  cepts for investigating truth, although we  assert that the precepts are referred either   * Note, that it is unfair to fix on a particular part  of the world in proof of what it was in the whole. States  and cities may advance themselves for a time by a  partial policy which keeps others backward : but the  policy will fail in the end. By a natural course of  things the advanced state will merge in the mass and  improve it : and thus the world will keep on advancing,  although the spectator, who contemplates only the  particular state, will think it is retrograding.     SECT, iil.]     165     to a false principle, or to no principle at all  fitted to unite them into one body of sys-  tematic instruction. The work lately referred  to *, fnrnishes, for instance, many excellent  precepts for avoiding errors in the use of  words, and for guarding against the snares of  sophistry; and if such precepts and such ex-  amples as it offers, distinct from the doctrine  of the syllogism, were industriously collected,  and brought forward in aid of the Logic  which all men learn and all men use, they  would be of inestimable value. A useful  system of Logic will guard our notions from  error not only while we think, but while we  are reasoned witht: for one chief way by  which truth enters the mind, is through the     * Viz, Whately's Logic.   + Our meaning will be understood ; but wc express  it by ii distinction which is grounded on no real dif-  ference. He who is reasoned with, if he understands  the ai^ument, is set a thinking ; and his agreeing or  disagreeing with the argument is the effect of his own  thoughts, however these may be set in motion, and  perhaps unreasonably influenced, by what he hears.     1S6.     QCHAP. II.     medium of language as employed by others :  and Logic should therefore arm us with all  possible means for coming at truth so offered,  through the various entanglements by which  the medium may be accompanied. Hence,  the various sophisms of speech accompanied  by their appropriate names, would still occupy  a place in such a Logic ; nay, for this purpose,  and for this alone, would the Aristotelian  doctrine of the syllogism deserve explanation ;  namely to understand how a conclusion drawn  from mere terms, may, as a conclusion from  them, be perfectly true and perfectly useless,  and thus to induce us to bottom all our  reasoning on things. — Having thus offered,  on the first of the questions proposed in Sect.  18, such observations in the affirmative as we  thought it required, we now proceed to the  second question.   22. That question was. Whether we ought  to consider the limits of Logic as extending  beyond the bounds proposed at the com-  mencement of this chapter : towards answering     SECT. 22.] ON LOGIC.     1G7     which, we may first inquire how far other  views of it extend. By the Scotch metaphy-  sicians, and generally in the schools of North  Britain, the word Logic seems to be so used  as to imply the cultivation of the powers of  the mind generally, correspondently with  M'atts's definition of tlie purpose of Logic,  namely, " the right use of reason." " I  have always been convinced," says DugaJd  Stewart*, " that it was a fundamental error  of Aristotle, to confine his views to reasoning  or the discursive faculty, instead of aiming at  the improvement of our nature in all its parts."  And he then goes on to mention the following  as among the subjects that ought to be con-  sidered in a just and comprehensive system  of Logic. " Association of ideas ; Imagina-  tion ; Imitation j the use of language as the   GREAT INSTRUMENT OP THOUGHT ; and the   artificial habits of judging imposed by the  principles and manners in whicli we have     * Fhilotiuphical Essays.  Chap. II.     Preliminary Disscrtatio     16s ON LOGIC. [|CHAP. 11.   been educated." * Now if the threeibld di-  vision of human knowledge is a just one,  which, in the Introduction of this work, was     his     * io the same purpose,  Philosophy of the Humat     n the second volume of  Mind, (Chap. III. Sect.     S.) he speaks thu^     ' The following, (which     mention by way of specimen,) seem to be among the  most powerful of the causes of our felse judgments.  1. The imperfections of language both as an instru-  ment of thought, and as a medium of philosophical  communication. 2. The difficulty in many of our  most important inquiries of ascertaining the facts on  which our reasonings are to proceed. 3. The partial  and narrow views, which, from want of information,  or some defect in our intellectual comprehension,  we are apt to take of subjects which are peculiarly  complicated in their details, or which are connected  by numerous relations with other questions equally  problematical. And lastly, (which is of all perhaps  the most copious source of speculative error) the pre-  judices which authority and fashion fortified by early  impressions and associations, create to warp our  opinions. To illustrate these and other circumstances  by which the judgment is apt to be misled in the  search of truth, and to point out the most effectual  means of guarding against them, would form a very  important article in a philosophical system of Logic,"     SECT. S2.]     169     borrowed from Locke,— namely into, it., the  knowledge of things tiiat are, — ii., of things  fitting to be rfonc, — and, Hi., of the means of  acquiring and improving both these branches  of knowledge;— it wUl at once appear that  all the subjects referred to in this enumeration  of Stewart's, except the fourth, which we print  in capitals, come under the denomination of  physica : — they are energies or tendencies of  the mind derived from nature, or habits  arising out of natural causes ; and they come  accordingly under the division of things ex-  isting in nature, which things, as they all  concern the mind, it is the business of the  Pliilosophy of the human mind to explortf:  but the fourth of the subjects mentioned in  the quotation from Stewart, viz •* the use of   LANGUAGE AS THE GREAT INSTRUMENT OF   THOUGHT," comes under the third of the  divisions laid down by Locke, and ought cer-  tainly to be distinguished from the other  subjects, because it is the means of becoming  acquainted with them : it is the instrument.     m     ON LOGIC.     [chap. II.     and they are among its objects. True, we  discover, as we proceed in the use of it, and  we are properly warned by those who have  used it before, that its efficacy is assisted or  impeded by extraneous causes, as well as by  defects in the instrument itself: similar dis-  coveries will be made, and similar warnings  must be given, in the practice of almost every  art: but these ought not to enter into the de-  finition of the art, although it will be proper  to bring them forward, incidentally, as we  open its rules. " A method of invigorating  and properly directing all the powers of the  mind is indeed," says Dr, Whately, " a most  magnificent object, but one which not only  does not fall under the province of Logic, but  cannot be accomplished by anyone science or  system that can even be conceived to exist.  The attempt to comprehend so wide a field is  no extension of science, but a mere verbal ge-  neralization, which leads only to vague and  barren declamation. In every pursuit, the  more precise aud definite our object, the more     SECT. 22.]     ON LOGIC.     171     likely we ai'e to obtain some valuable result j  if, like the Platonists, who sought after the  avTodyaSov, — the abstract idea of good, —  we pursue some specious but ill-defined  scheme of universal knowledge, we shall lose  the substance while grasping at a shadow, and  bewilder ourselves in empty generalities." *■  To these just remarks, we may add our ex-  pression of regret that Dugald Stewart never  had opportunity to do more than speak pro-  ^'^ectively of *' a just and comprehensive  system of Logic ;" " to prepare the way for  which, was," he says, " one of the main  objects he had in view when he first entered  upon his inquiries into the human mind."t  Had he himself completed such a design in-  stead of leaving it for others, we doubt not he  would have found the necessity of circura-  scribing Logic within the bounds we have  proposed, in order to give it existence as an     • Whately's Logic ; Introduction,  t Pliilos. Essays. Prelim. Diss. Chap. II.: in the  paragraph immediately following the last quotation.     fjtt ON LOGIC. [chap. U.   art distinct from the wide ocean of intellectual  philosophy.   23. But Dr. Whateiy, who deems, with  us, that every consideration of the mind con-  ducted without reference to its making use of  language as its instrument, lies out of the de-  partment of the teacher of Logic*, com-  pletely differs from us, as to the province of  the art. Of the question, " whether it is by  a process of reasoning that new truths are  brought to light," he maintains the negative t,  and consequently denies that investigation be-  longs to Logic. Afler what has been ad-  vanced in the former sections of this chapter,  we think it quite unnecessary to combat this  opinion here ; and as Dr. Whateiy concedes,  that " if a system could be devised to direct     • Dr. Whateiy defines Logic (Chap. II. Part I.  Sect. 2.) " the art of employing language properly for  the purpose of reasoning." But with him, reasoning  B argumentation.   t Whateiy "s Logic, Province of llcasoning, Chap.  II. Sect. 1.     ^     SECT. 23.] ON LOGIC. 173   the. mind in the progress of inveBtigation ", it  might be " allowed to bear the name of Lo-  gic, since it would not be worth while to con-  tend about a name " *; — as, moreover, we  propose to comprehend under Rhetoric all  that belongs to the proving of truth — that is,  convincing others of it after we have found it  ourselves ; — we might be satisfied with stating  that this is the distribution we choose to  adopt, and there let the matter end. Be-  lieving, however, that our reasons will shew  this distribution to be not only useful, but al-  most indispensable, we proceed to offer them.  24, And first, that, so far as we have  gone, the art we have described ought to be  called Logic, we think will hardly now be de-  nied: — for we have proved that from be-'  ginning to end, it is a process of reason, that  is to say, a process to reach an end by mediae  and we have shown that the media are     • Whalely't* Logic, Province of Jteasoiiing, Chap.  II. Sect. 4.     Wi     ON LOGIC.     [chap. II.     words, (Xo'yoi.) If the term Logic is not pro-  perly applied to such an art as this, we know  not where an instance can be found of pro-  priety in a name. But shall we include the of-  fice of proving truth under this name, as well  as that of investigating it ? We answer, no, for  these two reasons : first that the things them-  selves are difierent, and ought therefore to be  assigned to different departments ; since it is  one thing to find out a truth, and another to  put a different mind in a posture for finding it  out likewise : And, second, that persuasion by  means of language, which is the recognized  office of Rhetoric, is not so distinct from con-  viction by means of language, as to admit of  our saying, precisely, where one ends, and the  other begins. That common situation in life.  Video meUora proboque, deteriora sequor,  proves indeed there are degrees of conviction  which yield to persuasion, as thei'e are other  degrees which no persuasion can subdue : yet  perhaps we shall hereafter be able to show,  that such junctures do but exhibit one set of     SECT. 24.]     175     motives outweighing anol^ier, and that the ap-  plication of the term persuasion to the one set,  and of conviction to the other, is in many cases  arbitrary, rather than dictated by a corre-  spondent difference in the things. If, then, the  finding a truth, and the proving it to others,  ought to be assigned to different departments  of Sematology, why not, leaving the former to  Logic, consider the latter as appertaining to  Rhetoric, seeing that convincing is not always,  and on every subject, clearly distinguishable  from persuading, which latter is the acknow-  ledged province of Rhetoric ? Thus will ana-  ^5ii' uniformly belong to Logic, and synthesis  to Rhetoric. While we use language as the  medium for reaching further knowledge than  the notions (knowledge) we have already  gained, we shall be using it logically : when,  knowing all we intend to make known, we  employ it to put others in possession of the  same knowledge, we shall be using it rhe-  torically. As learners we are, according to  this distribution, to be deemed logicians }— .as     176     [chap, II.     teachers, rhetoricians. The two purposes are  quite distinct, though they are often con-  founded under the same name, reasoning ;  which sometimes means investigation, and  sometimes argumentation*, or a process with   • 111 spite of all we have said against taking up no-  tions from mere terms, (for " what's in a name ?") we  confeES a strong antipathy to the word argumentatmi.  It no sooner meets our eyes, than, fearing the approach  of some Docteur Pancrace, we instinctively put our  hands to our ears. " Voub voulez peut-etre savoir, si  la substance et Vaceident sont termes synonymes on  equivoques k I'egard de Tetre? Sganarelle. Point  du tout. Je... Pancrace. Si la lo^ que est un art, ou  une science.^ Sgan. Ce n'est pas cela. Je... Pancr.  Si elle a pour objet les trois operations de I'esprit, ou  la troieieme seulement ? Sgan. Non. Je... Poner. S'il  y a dix categories, ou s'il n'y en a qu'une ? Sgan.  Point. Je... Pancr. Si la conclusion est Vessence  du sylle^sme ? Sgan. Nenni. Je... Pancr. Si  fessence du bien est mise dans I'appetibilite, ou dans  la convenancc? Sgan. Non. Je... Pancr. Si le  bien se rcciproque avec la fin ? Sgan. He, non! Je...  Pancr. Si la fin nous pent emouvoir par son etre reel,  ou par son Stre intentionel ? Sgnn. Non, non, non,  non, non, dc par tons lea diables, non. (Moli&re's  Mariage Force.) We join in our friend Sganarelle'g     SECT. 24.] ON LOGIC. 177   a view to proof: and the confusion is pro-  moted by the circumstance, that the two pro-  cesses are often used in subservience to each  other. Thus, when a writer sits down to a  work of philosophical investigation, it is to be  expected that the general truths he designs to  prove, are already in his possession ; but he  has to seek the means of proving them. Now  in searching for these, it is not unlikely, that,  with regard to the detail, he will frequently  come to conclusions different from those he  was inclined to entertain, though the final re-  sult he had entertained may remain un-  changed. At one moment, therefore, he is a  logician, at another, a rhetorician. His reader,  on the other hand, is a logician throughout :  in following and weighing the arguments offer-  ed, he is an investigator of the truths which   deprecation, wishing to shun all argumentation, except  of that quiet kind which takes place when the talkers  on both sides are disposed to truth, ilot victory. If  the word conveyed to us the notion of so peaceable a  meeting, we should have no objection to it ; but we  have confessed our prejudice.   N     178 ON LOGIC. I^CHAP. II.   the other undertakes to prove. In this man-  ner may the same composition, accordingly  as it exercises the inquiring or the demon-  strating mind, be considered at one time with  reference to Logic, at another with reference  to Rhetoric. Still must it be admitted, that  to investigate and to prove are different  things ; and conceiving there is sufficient  ground for confining Logic to the former  office, we shall conclude our chapter as we  began it, by defining Logic to be the right  use of WORDS with a view to the investiga-  tion of truth.     CHAPTER III.  ON RHETORIC.     Non posse Oratorem esse nisi viriim bonum.   AKG, CAP. I. LIB. XII. QtriN. 1N3.     1. In the chapter just finished, it was shown  that the use of language as a Logical instru-  ment, entirely agrees with the theory of Gram-  mar we ascertained in the first chapter, and  that, on no other principles than those which  arise from that theory, can Logic be pro-  fitably studied. We have now to show that  the use of language as a Rhetorical instrument  agrees with the same theory, and that the  view of the art hence obtained, lays open its  true nature, and the proper basis for its rules.  2. The language of cries or ejaculations,  which in the first chapter we started with,  may be called the Rhetoric of nature. To  this succeeds the learning of artificial lan-  guage ; and the process, whether of invention  N 2     180 ON RHETORIC. [CHAP. III.   or of imitation, brings into being the Logic  described in the preceding chapter. For  whether we invent a language, or learn a lan-  guage already invented, (presuming it to be  the first language we learn,) we must learn,  (if we do not learn like parrots,) the things of  which language is significant. All words  whatever, not excepting even proper names *,  express notions (knowledge) obtained from  the observation and comparison of many par-  ticulars ; and singly and separately, each word  has reference to the particulars from which  the knowledge has been gained. But it is by  degrees we reach the knowledge of which  each single word is fitted to be the sign. We  begin by understanding those sentences, or  single words understood as sentences>, that  signify our most obvious affections and wants,  and which, taking the place of our natural  cries, retain the tone of those cries as far as  the articulate sounds they are united with  permit. In all cases, as a sentence expresses   * Vide Chap. II. Sect. 7- ad fincm.     SECT. 2.] ON RHETORIC. 181   a particular meaning in comparison with the  general terms of which it is composed^ the  hearer may be competent to the meaning of  the sentence, who is not competent to the  full meaning of the separate words. A cry,  a gesture, may deprecate evil, or supplicate  good ; and a sentence which takes the place  of, or accompanies that cry or gesture, will,  as a whole, be quickly interpreted. But the  speaker and the hearer must have made con-  siderable progress in the acquirement of know-  ledge by means of language, before the one  can put together, and the other can separate^  understand, such words as, ^^ A fellow  creature implores"; "A friend entreats *\   It is by frequently hearing the same word in  context with others, that a full knowledge of  its meaning is at length obtained * ; but this  implies that the several occasions on which it   * Consult, on this subject, Chapter 4th of Du-  gald Stewart's Essay " on the Tendency of some late  Philological Speculations,^ being the fifkh of bis " Phi-  losophical Essays^.     182 ON RHETOnic. [chap. hi.   is used, are observed and comjiared; it im-  plies, in short, a constant enlargement of our  knowledge by the use of language as an in-  strument to attain it.   3. But he who uses language as a logical,  will also use it, when need requires, as a rhe-  torical instrument. The Rhetoric of nature,  the inarticulate cries of the mere animal, he  will lay aside ; or at least he will employ them  (and he will then do so instinctively) only on  tliose occasions for which they are still best  suited, — for the expression of feelings re-  quiring immediate sympathy. On all other  occasions, he will use the Rhetoric by which  a mind endowed with knowledge, may expect  to influence minds that are similarly endowed ;  and our inquiry now is, how the effect is pro-  duced;— how, by means of words, (taking  words to be nothing else than our theory of  language has ascertained them to be,) — how,  by such means, we inform, convince, and  persuade.   4. According to our theory, wobds are to     SECT. 4.] ON RHETORIC. 183   be considered as having a double capacity ;  in the first, as expressing the speaker's actual  thought ; — ^in the second, as being the signs  of knowledge obtained by antecedent acts of  judgment, and deposited in the mind ; which  signs are fitted to be the means of reaching  further knowledge. Now, when we use lan-  guage as a rhetorical instrument, we use it,  or at least pretend to use it, in order to make  known our actual thought, — in order that  other minds should have that information, or  be enlightened by that conviction, which we  have reached. Could this be done by a single  indivisible word — could we realize the wish  of the poet —   Could I embody and unbosom now   That which is most within me ; could I wreak   My thoughts upon expression, and thus throw   Soul, heart, mind, passions, feelings, strong or weak.   All that I would have sought, and all I seek,   Bear, know, feel, and yet breathe, into One Word*   Were this instantaneous communication with-  ♦ Byron's Childe Harold, Canto III. Stanza 97-     184 ON RHETORIC. £CHAP. III.   in our power. Rhetoric would be a natural  faculty, not an art, and our inquiry into  its means of operation would be idle. But  getting beyond the occasions for which the  Rhetoric of nature is sufficient, and for which  those sentences are sufficient that serve the  most ordinary purposes of life, an instan-  taneous communication from mind to mind, is  impossible. The information, the conviction,  or the sensitive associations, which we have  wrought out by the exercise of our observing  and reasoning powers, can be given to another  mind only by giving it the means to work out  the same results for itself ; and, as a rhetorical  instrument, language is, in truth, much more  used to explore the minds of those who are  addressed, than to represent, by an expression  of correspondent unity, the thought of the  speaker ; — rather to put other minds into a  certain posture or train of thinking, than pre-  tending to convey at once what the speaker  thinks. Contrary as this doctrine will ap-  pe$ir to common opinion on the subject, a very     6ECT. 4.] ON RHETORIC. 185   little reflection will show that it must be true.  For a word can communicate to another mind  what is in the speaker's, only by having the  same meaning in the hearer^s : but if it have  the same meaning, then it signifies no more  than what the hearer knows already, or what  he has previously experienced. And this is  plainly the case with sentences (words) in  familiar use, which signify what all have at  times occasion to express, which are used  over and over again for their respective pur-  poses, and of which, while uttering or hearing  them, we do not attend to the meaning of the  separate words, but only to the meaning of  the whole expression *. Here, it is confessed,  the communication is made at once ; but then  it is a communication which the hearer is pre-  pared to receive, because he has himself used  the same expression for the same purpose.  What is to be done when the information or  the conviction is altogether strange to the  mind which is to receive it ? In this case the   ♦ Refer to Chap. I. Sect 19.     ON RHETOKIC. QCHAP. HI.     speaker will seek in vain, as in the first case,  for an expression previously familiar to the  hearer; and he will have to form an expres-  sion. But how shall he form it? As words  have the power of representing only what is  known on both sides, he must form it not  with signs of what is to be made known, but  of what is already known. In this way, he  may produce an expression — whether that  expression take the name of sentence, oration,  treatise, poem, &c. * — which, as a whole, de-  notes that which his mind has been labouring  to communicate — the information, the con-  viction, or the sensitive associations he is de-  sirous that others should entertain in common  with himself. The necessity of so protracted,  so artful a process, must be set down to the  hearer's account, not to the speaker's. The  latter is (or ought to be) in previous possession  of what he seeks to communicate — he has  been through the process, and reached the  result : but that result he cannot give at once  ' Compiirc Chap. I. Sect. 20.     SECT. 5.] ON RHETORIC. 187   and gratuitously to others : he can but lead  them to it, as he himself was led, by address-  ing what they already know or feel ; and his  skill in rhetoric will be the skill with which,  for this purpose, he explores their minds. It  will be a process of synthesis on his part, and  of analysis on theirs. He will form an ex-  pression out of WORDS which signify what  they already know, or what they have already  felt : and the separate understanding of these  on their part, will enable them to understand  his expression as a whole. This being the  theory of Rhetoric which grows out of our  theory of language, we now proceed to show  that the actual practice of every speaker, and  of every writer, is in accordance with it.   5. To begin with Description and Narra-  tion : — Is it not obvious, that, to procure in  another mind the idea of things unknown, we  proceed by raising the conception of those  that are known ? An object of sight which  the party addressed has never seen, we give  an idea of by allusions made iu various ways     188 ON RHETORIC. [CHAP. III.   to objects he has seen :— or if, being new as a  whole, it is made up of parts not new, we give  the idea of the whole by naming the parts,  and their manner of union. An unknown  sound, or combination of sounds, an unknown  taste, smell, or feel, is suggested to another  mind by a comparison, direct or indirect,  with a known sound, taste, smell, &c. As to  conceptions purely intellectual, it is a proof  how little one mind can directly represent or  open, itself to another, that, in the first in-  stance, such conceptions can be made known  not by words that directly stand for them, not  by comparisons with things of their own  nature, but only by comparisons with affec-  tions and effects outwardly perceptible; as  would at once be obvious in tracing to their  origin all words that relate to the faculties and  operations of the mind *'y although it is true   * Thus afdrnvs^ amma^ +*'%»», originally signify  wind or breath : ^vfiog /Mevog^ mens^ impetuosity ; in-  tellect is from inter and lego, I collect from among ;  perception and oonceptUm are from capio I take, — a     SECT. 5.] ON RHETORIC. 189   that these words at last become well under-  stood names, that at once suggest their re«  spective objects, without bringing up the ideas  of the objects of comparison that once in-  tervened. In narration we proceed by similar  means. We presume the hearer to be ac-  quainted with facts or events of the same  kind as that which is to be made known,  though not with the particular event ; for we  \x%Q generalievmSy i. e. terms expressing kinds  or sorts, in order to form every more par-  ticular expression. If the hearer should be  unacquainted with facts or events of die same  kind, the communicator then has recourse to   use of the verb still common in such phrases as ^^ I  take in with my eye,'' and, " I take your meaning ;''  judgment is from jus dicere ; understanding suggests  its own etymology ; refleadon implies a casting or  throwing back again; imagination is from imago^  an image or representation; to thinks according to  Home Tooke, is from thing ; — " Res-^k thing (he says)  gives us refyr I am thinged,'' i. e. operated upon by  things. These are etymologies suggested by authori-  ties universally accessible ; — the curious in this depart-  ment of learning would be able to add much more.     IdO ON RHETORIC. [CHAP. III.   circuitous comparisons. If nothing is pre-  viously known to wliich the action or event  can, however remotely, be compared, the  attempt to make it known must be as fruitless  as that of giving an idea of colours to one  bom blind, or of sounds to one born deaf*.   * Not without reason does the angel thus speak to  Adam in the Paradise Lost :  High matter thou enjoin'st me, O prime of men,   — and hard : for how shall I relate  To human sense the invisible exploits  Of warring spirits ?  And he proposes to overcome the difficulty in the only  way in which it can be concaved possible to be over-   — what surmounts the reach  Of human sense, I shall delineate so  By likening spiritual to corporal forms,  As may express them best.   Far. Lost. Book 5. 1. 5G3.  Still must the discourse of the Angel have been unin-  telli^ble to Adam : for the latter must be supposed  ignorant not only of the things to be illustrated, but  of far the greater part of the illustrations. There  was no keeping clear of this defect in the philosophy  of die jwem, if, in a poem, we arc to look for philoso-  phy. The discourse even of Adam and Eve, though     SECT. 6.]     ON RHETORIC.     191     6. Thus, then, when we make use of  words in order to inform, we produce the  effect by adapting them to what the hearer  already knows. In using words in order to  convince and persuade, we produce the effect  in the same way. But to convince, it is ne-  cessary to inform — to acquaint the hearer  either with something he did not know before,  or with something he did not attend to ; and  the information is called the argument * or  proof. Thus the information that "Plato was  a philosopher," is an argument or proof that  he is deserving of respect: and the clear  testimony that " a man has killed another  maliciously," proves that the perpetrator is  guilty of murder. But why do we account  the information in the respective instances an  argument or proof of the conclusion ? For   Iieautifully fiimple, is tilled with alluaions to things  which the least philosophy will teach us they could not  be acquainted with.   * The word argument is commonly used iii the  sense we here assign to it ; though it is likewise often  used with » more coniprelicnBivc meaning.     192 ON RHETORIC. [^CUAP. III.   no Other reason than this, that it is addressed  to a notion (knowledge) previously acquired  of what persons are deserving of respect, (in  the first instance,) and of what constitutes the  crime of murder, (in the second instance.)  Take away this previous knowledge, and the  information remains indeed, and may perhaps  be clearly understood, but in neither instance  can it lead the hearer to the conclusion, —  that is to say, it will not then be an argument  for the end in view : it will communicate,  perhaps, what it professes to make known, but  there the matter will end. In every process,  then, by which we propose to convince others  of a truth, there are three things implied or  expressed : i. that which we intend to prove  true, and which, if stated first, is called the  proposition, if last, the conclusion : ii. the in-  formation by which we try to prove it, and  which is accordingly called the argument or  pro of; iii. the previous notion (knowledge) to  which the information is addressed, and  which is frequently called the datum ; being     SECT. 6.] ON RHETORIC. 193   that which is presumed to be already known,  and therefore conceded or given by the person  reasoned with ; on account of which, and  solely on this account, the information is  offered in the capacity of an argument or  proof. Now, here we have the parts of a  syllogism, (though in reversed order, viz. the  conclusion, the minor, the major,) and this  may serve to show, without having recourse  to the Aristotelian doctrine of the comparison  of a middle with extremes, why the form of a  syllogism, where necessary, must always be a  forcible way of stating an argument. For  first we state that which our hearer cannot  but. concede j (major ;) then we state that  which he did not know or attend to, in such  a way that he must receive it on our testi-  mony, or admit as evident as soon as it is  attended toj (minor;) and these two being  admitted, they are found to contain what we  proposed to prove: which we then draw  from them without the possibility of a rational  contradiction; (conclusion.) For example;   o     194 ON RHETORIC. [CHAP, III.   our hearer knows by experience what persons  are deserving of respect: he knows, then,  that   ** Every philosopher is deserving of respect.^   We then remind him of the fact which he has  learned from history, that   " Plato is a philosopher :''   Hence on his own knowledge we advance  the undeniable conclusion,   " Plato is deserving of respect''   Is this conclusion at all fortified — is the  process which led to it explained — by shew-  ing that a comparison of the terms independ-  ently of the things, produces the proposition  which expresses it ? Both the hearer and the  speaker must have the kno'wledgevfYiicYi the first  two propositions refer to, or the conclusion can-  not be drawn for any rational end : and if they  have the knowledge, they have the conclusion  in that knowledge. In convincing the hearer,  the speaker does nothing but remind him  that he (the hearer) has the necessary know-     SECT. 6.] ON RHETORIC. 195   ledge ; and the syllogism, we admit, puts the  matter home in a very forcible way : but that  is all : another form of speaking will oflen do  equally well : for instance, " Plato who is a  philosopher is deserving of respect." Whether  the truth is stated in this way, or in the for-  mer way, or in any other way, the extract-  ing of a middle and extremes out of the ex*  pression, and demonstrating that these agree  or disagree, is, we repeat it, a puerile addition  to the process that has previously taken place.  Again, with regard to the other example at  the beginning of the section: — Our hearer  knows, (suppose him to be a juryman,) either  of his own knowledge, or by the definition  laid down by the judge, that   ^^ Maliciously killing a man is murder.''^   This is the datum, or major. He receives in  charge, i. e. he is informed that A. B. killed a  man maliciously, which is tantamount to  saying that   " What A. B. did, is killing a man maliciously.*"   o 2     196 ON RHETORIC. [CHAP. Ill,   This information is to be the argument or  minor by which the conclusion is to be esta-  blished; but the juryman must be made sure  of its truth, — he must know it, — before he  can receive it in this capacity : — well, he is  made sure of its truth : — must he then go to  Aristotle, and be taught to compare the  middle with the extremes, in order to pro-  nounce his verdict that   " What A. B. did, is murder:''   that is, he is guilty of murder? Will he be  MORE satisfied with his own verdict, if he is  able to do so ? Common sense pronounces,  no. Let us, then, for ever have done with  the Aristotelian Syllogism ; admitting, how-  ever, in favour of the form of expression, that  to express (i.) the datum, — (ii.) the inform-  ation which, because it is addressed to the da-  tum, is an argument,— and (iii.) the conclusion  from them — in three distinct propositions, is a  very forcible way of stating a truth which we  have reason to believe our hearer is prepared     SECT. 7-] ON RHETORIC. 197   to admit the moment it is so stated. But  the syllogism thus detached from the artifice  of comparing a middle with extremes, is only  one among the innumerable ways of express-  ing a truth, which the custom of language  permits, and is no more the invention of  Aristotle in particular, than any of those  other forms that might be used instead  of it *.   7. This brief notice of the syllogism in  addition to what was advanced in the last  chapter, occurs by the way : — ^the point we  had in hand, was, to show that in convincing  others by means of words, we adapt our words  to what they already know. And this must  be evident from what has preceded. For we  previously proved, that, in order to inform,   * Our observations on the syllogism are not meant  to call in question the intellectual capacity of the in-  ventor. For what we conceive to be a just estimate  of his merits, we refer to Dugald Stewart'^s Second  Vol. of the Philos. of the Human Mind, Chap. III.  Sect. 3., near the middle of the section.     198     ON RHETORIC. [CHAP. III.     we adapt our words to what our hearers al-  ready know ; and we have just shown that the  process of convincing them, is a process in  which we address some information to a pre-  existing notion. Let us now see how this  doctrine tallies with the terras of art which  are already in recognised use ; and, as occa-  sion may offer, let us inquire if there be any  difference, and what, between conviction and  persuasion.   8. That every argument used to influence  others, is considered to derive its efficacy  from some pre-existing notion, opinion, or rul-  ing motive, whether permanent or transitory,  in the hearer, is evident from the following  and similar expressions : argumentum ad Judi-  cium, by which we signify that our inform-  ation is addressed to such general principles of  judgment as mankind at large are guided by :  argumentum ad hominem, by which we imply  that we address those peculiar principles by  which the individual man is actuated. Again ;  argumentum ad vtrvcundiam, argumentum ad     SECT. 8.3     ON RHETORIC.     ignorantiam, argumentum ad Jidem, argumcn-  tum ad passiones, all imply arguments (infoim-  ation) addressed to some partial motives of  judgment and action ; and in all these, the  conclusion arising out of the reasoning has  the same validity, as far as regards the mere  act of reasoning : it is the difference of the  data that makes it of very different value. A  conclusion from an argument addressed to  principles which all men recognise, is obvious-  ly a conclusion of universal force; but one  which arises from an argument addressed to  peculiar principles, can of course be convinc-  ing only to such as admit those principles.  So likewise a conclusion which arises from the  reverence entertained for the author of the  principles professed ; — or which follows in the  hearer's mind from his limited notions, and  would not follow if he were better inlorra-  ed ;— or which follows because of his faith,  and would not follow, if he had not that  iaith J— or because his passions are previously  disposed, and would not follow, if they were     «00     ON RHETORIC.     [chap.     otherwise disposed: — in these and in similar  cases, the argument is valid, and therefore ef-  fective with respect to the minds for which it  is adapted, but addressed to other and more  general motives or knowledge, it may be no  argument at all *. Here, then, we may  perhaps see how the difference arises between  conviction and persuasion ; — mere persuasion  is conviction as far as it goes ; but it is con-  viction arising out of partial data : the person  persuaded is conscious that the reasoning  process itself is right, but he suspects —  perhaps more than suspects — tliat the data  which he has permitted his inclinations to lay   • Hence, what is Rhetoric at one tune and to one  set of auditors, may be none whatever at another time.  Who has not admired tlie Rhetoric of Marc Antony,  (the Hpeecb over Ciesar's body,) in Shakspeare's play  of Jnhua Caesar ? But why do we admire it F Is it  such Rhetoric as would persuade all people under the  circumstances supposed ? No. But it is just such  Rhetoiic as was fitted for the multitude under those  circumstances; and we admire the dramatist who so  completely suits the oration to the art of the speaker,  und the minds of those whom be has to operate upon.     SECT. 8.] ON RHETORIC. 201   down, are wrong: he perceives another con-  clusion from other and less suspicious data,  though he has not resolution enough to em-  brace it : so that the case we referred to in  the last chapter* as being so common in life,  Video meliora proboque^ deteriora sequor,  amounts to this, — that we are divided between  two conclusions, the one drawn from data  which we know to have the sanction of uni-  versal consent, the other from data supplied  by private motives. Thus, when Macbeth is  bunging in doubt between the suggestions of  duty and ambition t, the conclusion from each  source is reasonably drawn : but he is not  ignorant of the different value of the respec-  tive sources. He has nearly determined in  favour of the conclusion drawn from duty,  when his wife enters, who, by addressing con-  siderations (information, arguments,) to his  known sentiments of greatness and courageous   * Chap. II. Sect. 2+.   f Shakspcare's Macbetb, Act I. Scene 7-      JBOS ON RHETORIC. [^CHAP. 111.   daring, persuades him to murder Duncan and  seize the crown.   9. So much for the terms of art by which  we signify the quaUty of the arguments we  use, as depending on the known motives, or  information, or disposition, of the persons  addressed : which terms suit our theory so  well, that they seem to be invented for it.  Nest, for the terms by which the arguments  themselves are technically distinguished.  First, we have a distinction of them into Ex-  ternal and Internal. Now, according to our  theory, every argument consists of some in-  formation which we communicate to the per-  son reasoned with : — but this information  may be something that he could not possibly  have discovered by any consideration of the  subject itself J or it may be something that he  might have so discovered ; in which latter  case, our information will amount to nothing  more than making him aware of what he had  overlooked. The former, then, will be an ex-     »■]     ON RHETORIC.     temal argument or proof; the latter, an in-  temal argument. Of the former, the evidence  in a court of justice is an example ; as are al-  so proofs from history and other writings, and  irom the testimony of the senses. Of the lat-  ter kind, are all arguments from what are call-  ed the topica or loci communes : — for instance,  from the definition or conditions of a thing j  as when certain lines are inferred to be equal  to each other from their nature or conditions  as being radii of the same circle : — from  enumeration ; as when we prove that a whole  nation hates a man, by enumerating the  several ranks in it, who all do so : — from nota~  tion or etymology ; as when we infer that Lo-  gic has reference to the use of words in  reasoning, from its connexion with the Greek  Xt'yw I speak, and \6yoi a word :— from genus f  as when we prove that Plato is deserving of ■  respect, by showing that he is one of a getius  or kind that is deserving of respect : — from  species ; as when we infer the excellence of ^  virtue in general from that which we observe      eo*     ON RHETORIC.     [chap. lit.     in some particular act of virtue : — anil so like-  wise of the same kind, namely internal, are  aiguments from the other well known topics ;  (not to prolong the instances, which are easily  imagined ;) from cause, whether efficient, JiJial,  Jbrmal, or material; from adjuncts, antecedents,  consequences, contraries, opposiles, similitudeSy  dissimilitudes, things greater, less, or equal:  &c. The deriving of arguments from these  internal topics*, is nothing more, on the part  of the speaker, than turning a subject into  every point of view that may suggest a some-  thing relating to it, overlooked perhaps by  the hearer, and which, by being brought to  his notice, and addressed to his pre-existing  notions, may prove, or render probable, the  proposition in hand ; and according to the de-  gree of force which the argument carries, it is   • The reader needs not be reminded how largely  this subject of topics, (or places for finding the internal  or artiiicial proofs in contradiGtinction to the external or  artificial,) ia treated by the ancients : for instance, by  Aristotle, by Cicero, (vide the book called Topu-a,)  and by Quinctilian.     SECT, y.]     ON RHETORIC.     205     deemed an instrument of conviction or of  persuasion. An argument from defimlion ; — -  (for instance from the conditions of a problem  or theorem j as where lines are required to he  drawn which are to be radii of the same cir-  cle J ) which argument is addressed to a notion  assumed among the general conditions of the I  reasoning ; (for instance, that " a circle is suct]^ ]  a figure that all lines, (called radii,) drawn, j  from a certain point within it to the circum-  ference are equal " ;) — an argument so derived  and so addressed, is demonstrative of the pro-  position which it is brought to prove : (e. g^  that the lines are equal.) An argument froni[1  enumeration, — (for instance, from a statement 1  of the several ranks that are found in a n&- ]  tion,) addressed to a notion that the parta J  enumerated are all the parts, (for instance^ j  that the several ranks of people that hate A. j  B. comprise the whole nation,) is also de-  monstrative with respect to that notion ; but  if the enumeration should not comprehend all  the parts in the hearer's notion of the whole,      90Si ON RHETORIC. [CHAP. III.   or if the hearer should doubt whether his own  notion is sufficiently comprehensive, no ab-  solute conviction takes place. Still, the enu-  meration may induce belief, and will in such  case be said to persuade, though not to con-  vince. The same might be shown of the ar-  guments derived from all the other topics.  Entire conviction would follow from any of  them, if the hearer were fully satisfied both of  the truth of what is offered in the way of ar-  gument, and of the correctness of his own no-  tion to which the argument is addressed : but  greater or less degrees of doubt may accom-  pany each of these, and greater or less de-  grees of doubt will therefore attach to the  conclusions which flow from them. We may  moreover observe, that the truths a speaker  has in view, do not always stand in need of  demonstration : they are perhaps admitted al-  ready, but it may be that they do not suffici-  ently influence the hearer's sensibilities. The  object of an argument will then be, to awaken  those sensibilities, and with this effect its pur-     SECT. 9.] ON RHETORIC. 20?   pose wiU stop : as, for instance, when in or-  der to awaken sensibility to the frail nature  of man's existence, (not to demonstrate it,)  the speaker draws his argument from simili-  tude :   Ah ! few and full of sorrows are the days  Of mieerable man ! his life decays  Like that fair flower that with the sun's uprise  Its bud unfolds, and with the evening dies.   Here, the argument is obviously meant for  persuasion. There may, at the same time, be  an ultimate truth in view, which the speaker  designs to enforce when he has prepared the  mind for receiving it; and he will then employ  arguments of a different kind, and address  them to notions of universal dominion. — But  with regard to any of the arguments which,  in this brief review we have glanced at —  whether external or internal, whether demon-  strative, or only inducing belief, whether de-  signed to convince, or fitted but to per-  suade, — the process accords with the theory  assumed: — the speaker adapts words to know-     208 OM RHETOftlC. [chap. IU.   ledge the hearers have already attained, or  to feeliugs they have already experienced, in  order to conduct them to some discovery he  wishes them to make, or to some unexperienc-  ed train of thought conducive to such dis-  covery.   10. The assumption of this as the great  principle of the art, will, in the next place,  enable us to clear it from certain misdirected  charges to which it has always been liable.  The expedients which the orator employs,  the various tropes and figures of which his  discourse is made up, are apt to be looked  upon as means to dissemble and put a  gloss upon, rather than to discover his real  sentiments*. That, like all other useful   * We refer more especially to the following pas-  sage with which Locke concludes his Chapter ^^ on the  Abuse of Words ;^ being the 10th of his 3d book.  ^^ Since wit and &ncy find easier entertainment in the  world than dry truth and real knowledge, figurative  speeches and allusion in language will hardly be ad-  mitted as an imperfection or abuse of it. I confess  in discourses where we seek rather pleasure and de-     SECT. 10.] ON RHETORIC. 209   things, they ^re sometimes abused*, nobody   • E/ 3f, ort /jieyaKa jSxa\J/£(£v av b xi^f^^^°^ d^Uag  Tn roKzuTn ^uvifAEi tcHv Aoywv, touto re Jtoivov eo'ti Kara  ^ivruv Tuv ayaOav* Arist. Rhet. I. 1.     light than information and improvement, such orna-  ments as are borrowed from them can scarce pass for  faults. But yet if we would speak of things as they  are, we must allow that all the art of rhetoric, besides  order and clearness, all the artificial and figurative ap-  plication of words eloquence hath invented, are for  nothing else but to insinuate wrong ideas, move the  passions, and thereby mislead the judgment, and so  indeed are perfect cheats : and therefore however  laudable or allowable oratory may rehder them in ha-  rangues and popular addresses, they are certainly, in  all discourses that pretend to inform or instruct, wholly  to be avoided ; and where truth and knowledge are con-  cerned, cannot but be thought a great fault either of  the language or the person that makes use of them.  What, and how various they are, will be superfluous  here to notice ; the books of rhetoric which abound in  the world, will instruct those who want to be informed :  only I cannot but observe how little the preservation  and improvement of truth and knowledge is the care  and concern of mankind ; since the arts of fallacy are  endowed and preferred. It is evident how much men     210 ON RHETORIC. [CHAP. III.   will deny : but to consider them by their very  nature as instruments of deception, only  proves that the objector utterly misconceives  the relation between thought and language.  These expedients are, in fact, essential parts  of the original structure of language ; and  however they may sometimes serve the pur-  poses of falsehood, they are, on most occa-  sions, indispensable to the effective communi-  cation of truth. It is only by expedients  that mind can unfold itself to mind;— lan-  guage is made up of them ; there is no such  thing as an express and direct image of  thought. Let a man's mind be penetrated   love to deceive and be deceived, since rhetoric, that  powerftil instrument of error and deceit, has its esta-  blished professors, is publicly taught, and has always  been had in great reputation : and I doubt not but  it will be thought great boldness, if not brutality in me,  to have said thus much against it. Eloquence, like  the fair sex, has too prevailing beauties in it, to suf-  fer itself ever to be spoken against. And it is in vain  to find fault with those arts of deceiving, wherein men  find pleasure to be deceived.'*'     SECT. 10.3 ON RHETORIC. 211   with the clearest truth — let him burn to com-  luunicate the blessing to others ; — ^yet can he,  in no way, at once lay bare, nor can their  minds at once receive, the truth as he is con-  scious of it. He therefore makes use of ex-  pedients : — he conceals, perhaps, his final pur-  pose ; for the mind which is to be informed,  may not yet be ripe for it :— ^he has recourse  to every form of comparison, (allegory, simile,  metaphor*,) by which he may awaken pre-  disposing associations : — he changes one name  for another, (metonymy,) connected with  more agreeable, or more favourable associa-  tions : — he pretends to conceal what in fact  he declares ; — (apophasis ; — ) to pass by what   * In referring to these and other figures of speech,  it is impossible not to be reminded of Butler'^s distich,  that   All a rhetorician'^s rules   Teach nothing but to name his tools.   The fact is as the satirist states it. But then it is  something to a workman to have a name for his tools ;  for this implies that he can find them handily. — May  we add to our remark, that the world is scarcely yet   p2     212 ON RHETORIC. [CHAP. III.   in truth he reveals ; — (paraleipsis : — ) he in-  terrogates when he wants no answer ;— (ero-  tesis ; — ) exclaims, when to himself there can  be no sudden surprise;— (ecphonesis; — ) he  corrects an expression he designedly uttered ;  — (epanorthosis ; — ) he exaggerates ;— (hy-  perbole ; — ) he gathers a number of particu-  lars into one heap; — (synathroesmus ; — ) he  ascends step by step to his strongest position ;  — (climax; — ) he uses terms of praise in a  sense quite opposite to their meaning ; — (iro-  nia ; — ) he personifies that which has no life,  perhaps no sensible existence ; — (prosopo-  poeia ; — ) he imagines he sees what is not actu-  ally present ;— (hypotyposis ; — ) he calls upon   aware how much it owes to such men as Butler, Moliere,  Shakspeare, Pppe ;r-^men who joined to other rich gifts  of intellect, that of plain sound sense, which enabled  them at once to see, in their true light, the vanities and  absurdities of (misqalled) learningp But for the histo-  rian of Martinus Scriblerus, his predecessors and suc-  cessors, the world might still be under the dominion of  a set of solemn coxcombs, whose whole merit consisted  in making small matters seem big ones, and themselves  to appear wiser than their neighbours.     SECT. 10.] ON RHETORIC. 213   the living and the dead ; — (apostrophe : — ) all  these, and many more than these, are the ar-  tifices which the orator* employs ; but they  are artifices which belong essentially to lan-  guage ; nor are there other means, taking  them in their kind and not individually, by  which men can be effectually informedy or  perstuidedj or convinced. Could the prophet  at once have made the royal seducer of  Uriah's wife fully conscious of the sin he had  committed, he would not have approached  him with a parable t : that parable was the  means of opening his heart and understanding  to the true nature of his crime ; and it is a  proper instance of the principle on which all  eloquence proceeds. It is true, we do not   * We trust the reader scarcely needs to be remind-  ed, that the word Orator isused throughout this treatise,  in the comprehensive sense which includes all who  wield the implements of Eloquence. In modem times,  the influential orator is read not heard ; or if heard,  his hearers are few in number compared with his  readers.   t 2 Sam. 12.     214 ON RHETORIC. [CHAP. III.   now make use of parables fully drawn out ;  but all metaphorical expressions, all compa-  risons direct or indirect, are to the same pur-  pose ; namely, that of bringing the mind of  the hearer into a state or temper fitted for the  apprehension of truth. Nor, (we repeat,)  must it be thought that the means referred  to, (excepting some instances in bad taste,)  are ornaments superinduced on the plain mat-  ter of language, and capable of being detached  from it : they are the original texture of Ian-  guage, and that from which whatever is now  plain at first arose. All words are originally  tropes ; that is, expressions turned (for such is  the meaning of trope) from their first pur-  pose, and extended to others. Thus, when a  particular name is enlarged to a general one,  as our theory shows to have happened with  all words now general, the change in the first  instance was a trope. A trope ceases how-  ever to be one, when a word is fixed and re-  membered only in its acquired meaning ; and  in this way it is that all plain expressions have     SECT. 11.]     ON RHETORIC.     originated. In a mature language, a speaker  or writer may, therefore, if he pleases, avoid  figurative expressions. But the same neces-  sity, the same strong feelings, which originally  gave birth to language, will still produce new  figures, or lead the speaker to prefer those  already in use to plain expressions, if, by  the former, he can touch the chords, or awaken  the associations, that are linked with the truths  iie seeks to establish.   11. Our theory of language, and conse-  quent theory of Rhetoric, will, in the next  place, no longer leave us to wonder at an ef-  fect, which Dr. Campbell has laboured to  account for with much ingenuity; namely,  that nonsense so often escapes being detect-  ed both by the writer and the reader*. For  according to our theory, words have a sepa-  rate and a connected meaning, each of which  is distinct from the other. Now, suppose a  succession of words to have no connected     Chap. VII.      See Philosophy of Rhetoric, Vol. II. Book II.     216 ON RHETORIC. [CHAP, III.   meaning, which is as much as to say, suppose  them to be nonsense ; yet, in their separate  capacity, they will nevertheless stand for  things that have been known and felt ; and  if both the speaker and the hearer shbuld be  satisfied with the vague revival of this know-  ledge and of these feelings, they will neither  of them seek for, and consequently will not  detect the absence of an ulterior purpose.  The effect which is produced by words thus  used, (or rather misused,) extends no further  than that produced by instrumental music,  and is of the same kind. For no one will  pretend that a piece of niusic expresses, or can  express, independently of words, a series of ra-  tional propositions ; yet it awakens some sen-  timents or feelings of a suflSciently definite cha-  racter to occupy the mind agreeably. Now  perhaps it is not an unwarrantable libel on  one half of the reading world, if we affirm,  that they read poetry and other amusing  composition for no further end, and with no  further effect, than the pleasure of such vague     SECT. 11.] ON RHETORIC. 217   Sentiments or feelings as spring from music :  and to such readers it is of little moment  whether the words make sense or not. Ac-  cordingly, when composition like the follow-  ing is put before them^ which presents striking  though incongruous notions, in words gram-  matically united, agreeably jingled, and having  a connexion, probably, with certain sensitive  associations, they are liable to read on, not  only without feeling their taste shocked, but  perhaps with some pleasure.   Hark ! I hear the strain erratic  Dimly glance from pole to pole ;   Raptures sweet and dreams ecstatic,  Fire my everlasting soul.   Where is Cupid's crimson motion,   Billowy ecstasy of wo ?  Bear me straight, meandering ocean,   Where the stagnant torrents flow.   Blood in every vein is gushing,  Vixen vengeance lulls my heart ;   See, the Gorgon gang is rushing !  Never, never let us part *.   * " Rejected Addresses ;^ the particular example     S18     ON HHETORIC. [CHflP. III.     Nor is it in (pretended) poetry alone, that the  eflFect here alluded to tahes place. Bring to-  gether the rabble of a political party, and  place before them a favourite haranguer: — it  13 not by any means necessary that he should  make a speech which they understand, or even  himself: he has only to string, in plausible  order, the accustomed slang words of the  party, and to utter them with the usual fer-  vour ; the wonted huzzas will follow as a  matter of course, and fill each pause that the  speaker's art or necessity prescribes. And  BO likewise in an assembly of a different de-  scription, — the piously disposed congregation     above being in ridicule of Rosa Matilda's style. See  also Pope's " Song by a Person of Quality." The  reader whose taste is gratified by such composition as  is here caricatured, stands at the other extreme from  that mathematical reader, who returned Thomson's  Seasons to the lender with an expression of disgust,  that he had not been able to find a single thing proved  from the beginning to the end of the book. The  reader for whom the genuine poet writes, is equally  removed from each extreme.     SECT. 12.] ON RHETORIC. 219   of a conventicle : the good man whom they  are accustomed to hear has but to put to-  gether the words of familiar sound and evan-  gelical association — grace, and spirit, and  new light, regeneration and sanctification,  edification and glorification ; an inward call,  a wrestling with Satan, experience, new birth,  and the glory of the elect ; interweaving the  whole with unceasing repetitions of the sa-  cred name, accompanied by varied epithets of,  blessed, holy, and divine : and with no further  assistance than the appropriated tone and  frequent upturned eye, he will throw them  into a holy transport, and dismiss them, as  they will declare, comforted and edified.  This effect, which is apt to be attributed to  hypocrisy because the ordinary notions of  language suggest no cause for it, our theory  explains with no heavy scandal to the parties.   12. Concerning the elements of Rhetoric  ranged under the divisions of Invention and  Elocution, we have now made what remarks     220 ON RHETORIC. []CHAP. III.   our object required. There yet remains one  division, namely, Pronunciation *; which will,  however, scarcely furnish occasion for extend-  ing our observations ; since our theory is not  in any peculiar manner concerned with it.  As we started with the Rhetoric of nature,  namely, tone, looks, and gesture, so we are at   * Disposition and Memory are in general adde4  to these three. " Omnis oratoris vis ac facnltas,'*^  says Cicero, ^^ in quinque partes est distributa ; ut  deberet reperire primum, quid diceret; deinde in-  venta non solum ordire, sed etiam momento quodam  atque judicio dispensare atque componere ; tiun ea de-  nique vestire, atque omare oratione ; post, memoria  sepire; ad extremum, agere cum dignitate et venustate.^  De Orat. 1. 31. As to two of these divisions, we have  no occasion to notice them, because there is nothing  in our theory of language which requires them to be  viewed in a new or peculiar light : — We may take oc-  casion to observe, before' concluding the note, that the  modem use of the term Elocution, assigns it to sig-  nify what the ancients denoted by Pronunciation or  Action : and Dr. Whately sanctions this modem sense  by adopting it in his Rhetoric. We have used it  in the foregoing page in the ancient sense : ^^ quam  Graeci f^aa-iv vocant,^ says Quinctilian, ^^ Latine  dicimus Elocutionem.'*'* Ins. viii. 1.     SECT. 12,] ON RHETORIC. 221   once ready to admit that these may, and  ought to accompany the language of art ; —  that they ought not to be absent even from  the recollection of him who writes, lest  his style be deficient in vivacity. In union  with these parts of Pronunciation, is that ele-  ment of artificial oral speech called Empha-  sis ; and it will be to our purpose to observe,  how very inadequate are the common notions  of language to account for the actual practice  of emphasis, as it may be observed in English  speech. The common view of words that  make up a sentence, is, that they respectively  correspond to ideas that make up the thought :  and therefore, in a written sentence, if we  would know the emphatic word, we are de-  sired to consider which word expresses the  most important idea*. Thus, when Dr.   * To this end some teacher of elocution (elocution  in the modem sense) somewhere says : ^^ If, in every  assemblage of objects, some appear more worthy of no-  tice than others ; if, in every assemblage of ideas,  which arc pictures of those objects, the same difference     222 ON RHETORIC. [CHAP. III.   Johnson was asked how we ought to pro-  nounce the commandment, ** Thou shalt not  bear false witness against thy neighbour/* he  gave as his opinion that not should have the  emphasis, because it seemed the most im-  portant word to the whole sense. But Garrick  influenced by no assumed theory, pronounced  according to the practice of English speech,  ** Thou shalt-not bear," * &c. There is in fact  no other rule than custom in English speech  for the accenting of words in a sentence, any  more than there is for accenting syllables in a  word. A peculiar or referential meaning  may indeed disturb the usual accent of a   prevail, — it consequently must follow, that in every  assemblage of words, which are pictures of these ideas,  there must be some that claim the distinction called  emphasis.^ All this ingenious parallel, with Aristotle^s  authority to back it, we affirm to be purely visionary,  and we hope the reader by this time thinks as^ we do.  Yet is the passage in entire accordance with the no-  tions of language that commonly — nay, it should  seem, universally prevail.   * The story is somewhere related by BoswelL     SECT. 12.3 ON nHETOrtic. 223   word : for instance, the common accent of  the word for^ve, will be displaced if the  word is pronounced referentially to a word  that has a syllable in common ; as in saying  to give and loj'drgive. And just so will it be  in a sentence which is pronounced refer-  entially to an antecedent or a subsequent  sentence, either expressed or understood :  which would be the case, if we pronounced tie  ninth commandment in contradiction to one  who had said "Thou shaltbear false witness,"  &C., for then we should accent it in Johnson's  way, and say " Thou shalt n6t bear," &c.  Now this is what is properly called emphasis,  namely, some peculiar way of accenting a  sentence in order to give it a referential mean-  ing. A sentence pronounced to have a plain  meaning has its customary accents, but no  emphasis. The commonest example will be  the best ; and therefore we will quote one  that may be found in every book in which  emphasis is treated of: " Do you ride to  town to-day ?" If this is pronounced with-      •294 ON RHETORIC. [CHAP. HI.   out allusive meaning, ride, town, and day,  are equally accented by the custom of the  language, and there Is no emphasis properly  so called : which, by the way, is a pronun-  ciation of the sentence that teachers of read-  ing, in their search after its possible oblique  meanings, forget to tell us of. Suppose we  give an emphasis to ride, then lide-to-toivn-to  day will be allusive to ■wdlk-to-town-to-day, as  we might accent the word intrinsical in the  mauner marked with a reference to the word  Extrinsical, although the plain accentuation is  intrinsical. So again to-loTvn-lo-day is allusive  to the-country-to-day, and to-town-to-ddy is al-  lusive to to-town-to-m6rrow ; as the word  powerless might be accented on the last syl-  lable with a view to poweiiful. That the ac-  tual practice of emphasis corresponds with  this account, the reader may satisfy himself  by observing the conversation of the well-  bred, — not their reading, for that is oflen  conducted on mistaken principles : — and we  scarcely need point out how completely this     SECT. 12.] ON RHETOIUC. 2@5   practice accords with our theory of language.  For with us, a sentence is a word, not more  resolvabie into parts that constitute its whole  meaning, than a word made up of syllables ;  and as with regard to a word of the latter de-  scription, the accent is determined to one syl-  lable by custom, but is disturbed and placed  on another syllable in making allusion to  another word having syllables in common ;  so with regard to a sentence (word) made up  of words, the accents are likewise determined  to certain words that usually bear Ihem, but  these accents are disturbed and placed on  other words in making allusion to a meaning  which has, orwhich, if expressed, would have,  words in common. And here, with this new  kind of proof in favour of our theory, and  with the last subject usually treated of in  Rhetoric, we might stop the hand that has  traced this OutHne. But there remain a few  remarks that could not be introduced earlier,  for which the patience of the reader is en-  treated a little longer.      226 ON RHETORIC. [CHAP. III.   13. We may take the liberty in the first  place to observe, that, with regard to the  materials of Sematology which have been con-  sidered, our theory leaves them what they  were : it pretends only to show the true basis  on which they stand, and that the learned  distribution of them, is not that which accords  with the actual practice of mankind. Suppose  then, (if we may suppose so much,) that our  Grammars, our Books of Logic, and our In-  stitutes of Rhetoric, are to be altered in con-  formity with the views which have been  opened, the changes will not affect the detail,  but the general preliminary doctrine, and the  subsequent arrangement. As to doctrine,  the changes will mostly consist of omissions.  In Grammar, if we omit the common de-  finitions of the parts of speech *, and allow   * God help the poor children that are set to learn  these, and other of the definitions in elementary  grammars, particularly English grammars; for the  Latin ones are a little more sensible. That jumble of  a grammar that has the name of a Lindley Mturay in  the title page, after defining a verb to be ^^ a wend     I     SF.CT. 13.J     ON RHETORIC.     227     the tyro to learn what they are by the parsing  of sentences — that is, to ascend from par-   ihat Bignifiea to be, to do, or to suffer," {as if no other  part of speech signified to be, to do, or to suffer,) —  after saying what is true enough, but cannot be under-  stood by a child till he has practically discovered it,  that " common names stand for kinds containing many  sorts, or for sorts containing many individuals under  them;" — with many like things, picked up from  Lowth and others, equally fitted for the instruction of  young minds; condescends to give a few plain di-  rections for knowing the parts of speech, such as the  tyro is likely to understand: but the author, as if  ashamed of having been intelligible, remarks that  " the observations wliich have been made to aid  learners in distinguishing the parts of speech from one  another, may afford them some small assistance ; but  it will certainly be mucli more instructive to distinguish  them by the definitions, and an accurate knowledge of  their nature" Now the observations referred to, are,  in fact, the only passages calculated to give a just un-  derstanding of the parts of speech ; the definitions  wliich the writer enhances, being founded in an es-  sentially wrong notion of the nature of grammar. It  is speaking to the purpose to tell the tyro that " a  substantive may be distinguished by its taking an  article before it, or by its making sense of itself;"^ that,  " an adjective may be known by its making sense with  q2     gSS ON RHETORIC. [CHAP. III.   ticulars to generals instead of descending  from generals to particulars, — there la nothing   the wortl thing, or any particular Gubstantive ;" that,  " a verb may be diBtinguishcd by its making sense  with any of the personal pronoiuiB ;" that, " a preposi-  tion may be known by its admitting after it a personal  pronoun in the objective case ;" and so forth. These  are not only plain directions for the purpose professed,  but they suggest the real differences among the parts  of speech; and if the compiler had condescended  throughout his book (or books, for there are appen-  dages) to adapt his explanations, in the same manner,  to the minds of those who were to be taught, he would  have avoided the errors of doctrine which he always  runs into when be attempts to give, what as the author  of an elementary grammar he has never any buaiiiesa  to give, namely a philosophical or general principle.  Moreover, in the arrangement of his materials, he  seems incapable of, ot at least is inattentive to, the  clearest and most necessary distinctions. Thus, (to  take at random two examples from liis book of ex-  ercises,) he gives the following as instances of bad  grammar : " Ambition is so insatiable, that it will  make any sacrifices to attain its objects." (12mo. edit,  p. 128.) " When so good a man as Socrates fell a  victim to the madness of the people, truth, virtue, re-  ligion, fell with him." (Ibid 116.) The former of  these sentences exemplifies the Logical fault, non-     SECT. 13.] ON RHETORIC. 229   in what remains that can be objected to : the  declining of nouns, the conjugatiiig of verbs,   scquitur, and the latter will advantageouBly receive  the Rheimcal ornament polysyndeton : but to give  them as instanccB of defective Grammar, b to blind  the learner to the nature of the art he is studying. —  The grammatical works wc are referring to, seem,  from the number of editions they have gone through,  to be in very general iise, or we should not have  deemed them worth so long a note. \Ve pass to a  remark on another grammatical work of very different  character and value, the Greek grammar of Matthise.  This work has justly won the approbation of the  learned throughout the world; but we conceive the  praise belongs to its elaborate detail, and not to such  principles as the following. " Every proposition, even  the simplest, must contain two principal ideas, namely  that of the Subject a thing or person, of which any  thing is asserted in the proposition, and that of the  I'redicate, that which is asserted of that person or  thing." (Matth. Gr. § 293.) To state our objections  to tliis passage is difficult, because we do not know how  the author or translator may define a propositic»i, or  what they may mean by the principal ideas in it.  Perhaps they may consider no expression a proposition  which does not consist of a subject and predicate. Wc  deny that, from the nature of the thought, any commu*  nication requires these grammatical parts, {they are     A     380 ON RHETOKIC. [CIIAP. III.   and the other business of the grammar-scliool,  we deem, as it has always been deemed, in-  dispensable. In Logic, if we omit ail that is  taught concerning ideas independently of  words ; if we omit what ia taught concerning  the two operations of the mind, Perception  and Judgment distinct from Reasoning, not  because those operations do not take place,  but because every single abstract word fully  understood, (and Logic begins with words,)  expresses a conclusion from a rational process  as efTectually as a syllogism ; and if we further  omit (and the omission is important) whatever  is peculiar to Aristotelian Logic ; — all that  remains will, on the principles we have had  before us, be essentially useful to the learner ;  namely, the precepts for accurate definition ;  the precepts against the assumption of un-  warranted premises j the precepts for guarding  against the false conclusions to which we are     merely g^rammalical,) though the necessities of lan-  guage in general prescribe them. See Chap. I. SecL  25. ; about the middle of the Section.     SECT. 13.]     ON RHETORIC.     231     liable when we reason tvith words, and not  merely by means of words; the precepts for  guarding against being led away by true con-  clusions, when there may be conclusions like-  wise true and more important from other  data ; which data, with their conclusions, are,  kept out of sight by the art of the speaker, or .  the blindness of the inquirer*. In Rhetoric,  there is less to be omitted than in the other  branches ; but in this department, the general  views we have opened are important, because  they exhibit the art in connexion with a great  and worthy end; an end which, it should seem>  has not always been thought essential to it.     * We mean to say, that the7na(e)'taZsof acomplete  budy of ioEtructioD ia Logic already exist in Literature ;  but tliey esisE not in any one system. They are more-  over BO mingled with what is erroneous hi doctrine, that  the good is difficult to reach, without imbibing a great  many wrong notions that frustrate the practical benefit  How can it be otherwise, if what we have endeavoured  to prove, is true, that the principle of the Logic which  all men use and all men operate witli, has never yet  been cxpIaiRvd ?      j^P£ ON RHETORIC. [CHAP. III.   For as Rhetoric is an instrumental art, we  are told that it ought to be considered ab-  stractedly from the ends which the speaker  or writer may propose in using it j and  Quinctilian who insists that the Orator, (that  is, of course, the consummate orator,) must be  a virtuous man, lias been classed with those  whom atraihevffla, and aXai^ovela have betrayed  ioto a wrong estimate of the art*. As we  think the good old Roman schoolmaster is  not quite beside the mark in his notion on  this point, we propose to inquire wliether  the placing of Rhetoric on the basis we have  ascertained, does not lead to the position he  so stoutly maintains. Now, the immediate  basis of Rhetoric is Logic ; and our remarks  will therefore begin with the latter.   14. Logic as well as Rhetoric is an in-  strumental art ; but if our definition is correct,  it is an instrument for the discovery of truth,  and it is then only perfect as an instrument  when it is completely adapted to that end.  • See Whately's Rhetoric : Introductiun.     SECT. 14.] ON RHETORIC.     233      A great and worthy end is therefore essential  to Logic ; and a correspondent effect will  appear in those who have made a skilful use  of it. But the Logic we speak of, is that  which is applied to things, namely to Physicot  and Practica *; that is to say, which is em-  ployed to ascertain the constitution of the  world in which we Uve, and of ourselves who  live in it, and thence to deduce what we  ought to do: — but the examination of the  world, and of ourselves, and of our duties, is  the examination of particulars ; and our Logic  has recourse to universals for no other purpose  than to understand particulars the better. If  there is a Logic, which, resting in universals,  confers the power of talking learnedly and  wisely, yet leaves a man to act the part of an  Ignoramus and a fool in the commonest  concerns of life, this is not the Logic we have  had in view. There is indeed a learned ig-  norance, aa there is an ignorance from want  of learning ; there is also an ignorance from  * Cumparc ihc Intioduction.     m»     ON RHETORIC.     [chap. hi.     natural incapacity, and an ignorance from  superinduced insanity ; by any one of wliich  tbe mind may be prevented from reaching  truth. Not that in any case whatever the  reasoning process is wrong ; but if the  reasoning proceeds on wrong or insufficient  premises, which it will in any of these cases,  the conclusion will of course be wrong. Some  one has said that " the difference between a  madman and a fool is, that the former reasons  justly from false data, and the latter erro-  neously from just data." This is incorrectly  said : — the idiot who walks into the water  because he knows no better, is incapable of  the just datum, and therefore cannot be said  to reason from it : if he knew the datum,  namely that the water would drown him, he  would not walk into it ; but he does not  know this, and therefore he walks into it : in  doing which, he reasons, so far as his know-  ledge goes, as justly as the madman, who  walks into it because his disturbed fancy  makes him take it for a garden. Wlien the     SECT. 14.] ON RHETORIC. 235   road to truth is blocked up by either of these  two causes, namely irabeciUty or insanity.  Logic can do nothing ; but ignorance whether  from wrong learning or from want of learning,  is to be removed by the appUcation of ge-  nuine Logic to P/it/ska and Praclica. Still,  independently of tlie toil to be encountered,  there are obstructions and delusions which  are liable to turn the most ardent inquirer out  of the path. There may not be natural im-  becility, nor permanent insanity ; yet there  may be an habitual incapacity of judgment  from the influence of prejudice, and aa  occasional insanity of judgment from the in-  fluence of passion. But among other things  we learn in Pki/sica, these facts are to be  reckoned ; and the precepts which warn us of  them, are among the most important of those  which belong to Praclica. In the mean  time, that we may be induced to persevere in  the search after truth, till our real interests  become so plain that we cannot but embrace  them, we are not permitted to feel at ease     ^6 ON HHETOItlC. [CHAP. III.   under the mists which passion and prejudice  create. The fool and the madman to whom  mists are reaUties, are satisfied in their judg-  ments; but it is not so with those who see  dimly through the fog, and suspect there may  be better paths than those they are pursuing.  This suspicion, as light breaks in, may at last  become conviction, strong enough to subdue  even the habit or inclination by which a  wrong path is made easy, and a departure  from it difficult. True, indeed, such over-  powering conviction may not reacii the ma-  jority of mankind at present: they may be  compelled, as heretofore, to wear out life in  struggles between right and wrong, between  inclination and duty, between future good  and present solicitation : but are we forbidden  to hope, for future generations, a gradual  alleviation of so painful a conflict, in propor-  tion as what is good and what is evil shall be  made plainer to the eye of reason • P At least   > * All vice is ignorance or habit. Who would not  take the best way of being happy, if he knew it — that     SECT. H.] ON HHETORIC.     S37     may we affirm, that all learniag has, or ought  to have, this consummation in view.   is, knev it to conviction — and his habits did not prevent  him ? But he may discover the best way when hia  bahitE are fixed; as a miEerable dnmkard, who drinks  on to escape from utter dcepair, sees with bitter regrel  the happiness of a sober life. With a common notion  of learning and ignorance, an objector will demur to  our statement ; but such an objectot should be told,  that a man may have run the circle of the sciences aa  they are commonly taught, and yet remain in ignorance  of what is most important to be known. This is s  truth which not only Christian teachers, but the wise  among the heathen inculcate. In that admirable relic  of Socratic philosophy, £;EBHT02 niNAH, there  are, among the personifications, two that bear the  names of naiitia and "Htuimaihla, (Learning and  Counterfeit-learning,) by the latter of which is ligured  all that, independently of the knowledge which makes I  men permanently happy, passes under the name of I  learning. Now, in that knowledge which alone ia |  valuable, a man cannot be called learned, whose coik  viction is not strong enough to determine his practice.  The thirsty wight Tiho, in a state of profuse perspira*  tion, calls for a glass of iced-water, may know there is  danger in the draught : but if his knowledge is not  strong enough to prevent the act, what is its value ?—  at the moment, it is even worse than useless ; since      JiJ     SS8 ON RHETOIIIC. [chap. III.   15. Such then is the aim and scope of Lo-  gic in relation to Physica and Pracika : it is   may be sufficient to disquiet the luxury of the draught,  though not sufficient to subdue the desire for it.  When Macbeth, (for the case is not dissimilar,)  resolves to gratify his ambition, he is not ignorant of  the danger he runs, and the secure happiness he leaves  behind him ; but he is so far ignorant as to prefer the  phantom of happiness to the reality. Yet he is not so  ignorant as his wife, and he reaps, in consequence, less  immediate gratification. Having once held the balance,  with some impartiality, between right and wrong, he is  incapable, even for a moment, of being a triumphant  villain. The crooked-baek Richard, (for having begun  our examples with Shakspeare, we will continue with  him,) is not so distracted by divided data. " Securely  privileged," says Mr. Foster, " from all interference of  doubt that can linger, or hiunanity that can soften, or  timidity that can shrink, he advances with a grim con-  centrated constancy through scene after scene of  atrocity, still fiilfilling his vow to ' cut his way through  with a bloody ase.' He does not waver while he  pursues his object, nor relent when he seizes it."  (Essays on Decision of Character, &c.) Yet both he  and Macbeth's wife at length get nervous in their  sleep : for so it is, that if one scruple of conscience lurk  in the soul, it will produce its effect sooner or later;  and tliat effect will begin when the bodily powers are     F     SECT. 15.] ON RHETORIC. ^Q   the means of discovering truth in botli these  departments. Now we assume, that the pro-  weakest; and as body and mind have a mutual in-  fluence, the former -will sicken and perpetuate the  horrors of the latter, unless, as with Richard, a violent  death intervene. The three wretches vc have thus  far referred to, have this in common, that they do not  embrace vice for its own sake, but as a means of reaching  the phantom of happiness that dances before them.  But there is a state of vice brought on by habit, in  which a man finds a pleasure in doing evil, and is in-  capable of any other pleasure. lago is our example —  a character which, it is to be feared, is by no means  out of life. Imagine a shrewd and selfish child per-  mitted from infancy to create for himself a satis-  faction in the disquietude of others — a little worrier of  defenceless creatures— a petty tyrant indulged in his  worst caprices ; — imagine such a one, as he grows up,  placed where his habits cannot be indulged but in  secret, and where those around him are such, that he  must, in his own mind, either hate them, or hate  himself: imagine all this, and lago will appear too  possible a character. Some critics have objected, that  there is no sufficient motive for the mischief he brings  on Othello, Desdemona, and Cassio. Can there be, to  Aim, a stronger motive, than that they arc noble-  minded, benevolent, and happy, and tacitly remind  him, at every instant, that he is in all respects a      J     240 ON RHETORIC. [cHAP. IIF.   per business of Rhetoric is to make truth  known when found j which assumption, if ad-  mitted, would at once establish our position ;  for to suppose a consummate orator would, in  such case, be to suppose one who is too fully  possessed of truth not to be led by it himself,  while acting as a guide to others. After ad-  mitting the assumption, it would signify little   ■wretch? He knows and bitterly feels, tliat each  " hath a daily beauty in his life that makes him ugly-"  The only pleasure which habit has given him, in lieu  of those of which it has made him incapable, is, to  torture the beings that wound his self-love to the quick,  and to destroy the happiness he cannot partake in.  Such is the power of habit. Though the means, when  properly applied, of putting a human being in train to  become an angel, yet added to, and encouraging the  tendencies of his uninstructed nature, it will render him,  prematurely, a fiend. lago is utterly depraved — a be-  ing incapable of Paradise if placed in it — more odious  tlian Milton has been able to depict even Satan him-  self; for that majestic bdng, (the hero of the poem as  Drydeu truly says he is,) never appears " less than  arcliangel ruined. " The " demi-devil " of the dra-  matist, excels, in mental deformity, what the epic muse  has been able to conceive of " the author of all evil. "     SECT. 15.] ON RHETORIC.     241     to object the actual characters of those who  speak and write ; for they may be pretenders  in Rhetoric j or their advance in it, though  real, may be very inconsiderable toward the  perfection we are supposing. But it may be  said that the assumption begs the question,  and leaves us still to show that the office of ■  leading men to truth is essential to Rhetoric,  in contradiction to those who view it as a mere  instrument equally fitted for the purposes of  truth and falsehood. Now, it must be con-  fessed, with regard to the means employed in  Rhetoric, that they frequently seem adapted  to the prejudices of men, — to meet rather than  to oppose their ignorance and their passions.  And if there were any way of conveying truth  at once into minds unfitted to receive it *, the     * It is a comiuoii thing to say of a person, that he  vtiU not be convinced. The fact generally stands  thus : we use arguments that convince ourselves, and  presume they are fitted to convince him, not knowing  or not observing, that all argument derives its force  &om the previous knowledge in the mind to which it  is addressed ; and that our hearer may have been so        343 ON RHETORIC. [CMAP. III.   use of such means would be conclusive against  an honest purpose in the speaker. But the  instantaneous communication of truth, is, un-  der most circumstances, impossible ; and there-  fore we may next ask, what interest a writer  or speaker can have in an ultimate purpose to  deceive. The answer will be, — to serve one  or other of those partial purposes, of which  the common business of life, whether we look  into its private circles, or into the forum or  senate house, furnishes hourly examples. But  may we not describe all this as a conflict, in   educated as to render convicUon impoBsible by iuch  arguments as we offer him. Suppose, however, it be  true, that our hearer mill not be convinced, — thai is  to say, does not wish to be convinced, because his par-  ty perhaps, or his profession, or the career (be it what  it may) into which he has entered, does not agree witli  what is sought to be established : let us in candour  consider in such a case what a vantage ground we oc-  cupy, inasmuch as we see our own interest, temporal  or eterual, coupled with the proposition in view ; and  let us condescend, by the argumeittum ad homhiem,  to give him a similar advant^e, before we expect his  conviction from the argumentum ad judicium.     aECT. 16.] ON RHETORIC. 243   which each is eager to show just so much  truth as suits the present purpose, and to veil  the rest? And will not the whole of truth be  shown in this manner, as far at least as men  have discovered it, although not shown at  once ? Of these skirmishers that use the arms  ufiensive and defensive of the art, each takes  credit for a certain degree of skill j but among  them all, which is thg Orator? Is it not he  who soars above partial views and partial pur-  poses, who unites into one comprehensive  whole what others advocate in parts, who  teaches men to postpone petty for greater ad-  vantages, and to seek the welfare of the indi-  vidual in the happiness of the kind ? If, then,  the palm of eloquence is permanently his alone,  who contends for it in this manner, our chain  of argument will not want many links before  we reach the conclusion, that to undertake  the art on a valid principle, we must con-  sider its purpose to be that of leading men  to truth.   16. A Rhetoric growing out of the Logic     i     344 ON HHETOltrC. [^CH AP. HI   of Aristotle *, which, as we have seen, is the  art of reasoning mlh words, and not merely  by means of words, may indeed well be sus-  pected as a specious and delusive art. Aim-  ing at plausibility alone, it gives the power of  talking largely without requiring the know-  ledge which grows up Irom experience in  particulars ; and thus we have statesmen,  who, if we listen to them, are capable of setting  the world in order, but know not how to re-  gulate their households ; we have financiers  ready to accept the control of a nation's     • Aristotle's own treatise on Rhetoric is a work  completely to its purpose ; that is to say, fitted to make  men prevailing speakers at the time in wliich he wrote,  by exhibiting comprehensively the bearings of the ques-  tions they would have to discuss, and the various kinds  of persons they would have to influence. It is indeed  remarkable how little Aristotle's other works are of a  piece with his Logic ; nor is it without some show of  reason that Dugald Stewart supposes he was aware of  its empty pretensions, and was too wise to be deceived  by it himself, though lie chose to impose it on others.  Sec Vol. II. of the Philosophy of the Human Mind,  Chap, III. Sect. 3.     SECT. 16.] ON RHETOUIC.     245     wealth, that have never learaed to manage  their own estates; we have lawyers, whom  the simplest questions of right and wrong  would be sufficient to pei-ples * ; and priests  who, once a week, discourse " in good set  terms " to well dressed congregations, of vir-  tue and of vice, of this world and the next j but  who would be incapable of oifering, from their  own stores, a single argument fitted to deter  a plain thinking, ignorant man from vice, or  to stop the commission of a specific offence  by remonstrance adapted to the case. This  specious eloquence, however, like the Logic  from which it springs, has almost lost its re-  putation and influence: we now require from  speakers and writers more substantial recom-  mendations than the power of dwelling on  vague generalities ; and in proportion as   • But perhaps, with regard to lawyers, we are  requiring knowledge, which, as matters stand, would  be an incumbrance to them. A special pleader may  Bay, " what have I to do with simple right and wrong ?  My business is to see how the letter of the law can be  applied or evaded."     Mfi     ON RHETORIC. [CHAP. III.     genuine Logic enlarges the empire of truth,  will the necessity appear of seeking in an en-  lightened mind, and a heart kindled by active  philanthropy, for the true springs of elo-  quence. Thus will ambition be brought to  side with virtue} because there will be no  way of winning distinction, but by cultivating  the powers of language in subservience to  that knowledge, which gives a man the de-  sire and the faculty of beiug useful to others,  and governing himself.   17. To conclude ; — the theory which, in  this treatise, we have endeavoured to establiah  is this, — that we come at all our knowledge  by the use of media, which media are, chiefly,  words; and that, as the words procure the  notions, the notions exist not antecedently to  language : —that when, by these means, we  have gained knowledge, and try, by similar  means, to communicate it to others, we do  not, while the process is going on, represent  our own thoughts, but we set their minds a  thinking iu a particular train ; that our own     SECT. 17>3 ON RHETORIC.     247     thought 13 represented by nothing short of  the completely formed word, whose parts, if  any or all of them are separately dwelt upon,  are not parts of our thought, but signs of  knowledge which we and our hearers possess  in common, and which, by bringing their  minds into a particular attitude, enables them  to conceive our thought, when the whde  WORD that expresses it, is formed : — that i§  before this word is formed, there are parts by  which something is Communicated not known  before, yet, being communicated, it is still  but a part of the means toward knowing  something not yet communicated, and stiU,  therefore, the principle holds good, that we  are adding part to part of the whole word  which is to express something not yet com-  municated ; which word, even though it ex-  tend to an oration, a treatise, a poem, &c., is  as completely indivisible with respect to the  meaning conveyed by it as a whole, as is a  word which consists only of a single syllable,  or a single sound. If this doctrine truly de-  scribes the nature of the connexion between     248 ON RHETORIC. [^CHAP. III.   thought and language, we claim for it the  merit of a discovery, because the common  theory, that is, the theory which men are  presumed to act upon, and to which all pre-  ceptive works are adapted, — not the theory  which, unawares, they really act upon, — ex-  hibits that connexion in a very different light.  And, as a discovery, we are the more dis-  posed to urge attention to it, because our  soundest metaphysicians have expressed them-  selves as if there 'ooas something to be dis-  covered as regards the connexion we speak of,  before a system of Logic could be establisiied  on a just foundation. Locke says that when  he first began his discourse on the Under-  standing, and a good while after, he thought  that no consideration of language was at all  necessary to it. At the end of his second  book, he discovers, however, so close a con-  nexion between words and knowledge, that  he is obliged to alter his first plan ; and having  reached his concluding chapter, he speaks as  if he still felt that he had not yet ascertained  the full extent to which language is an instru-     SECT. 170 ^^ RHETORIC.     249     raent of reason. Dugald Stewart, too, from  whom, in the conclusion of our first chapter,  we quoted a passage which entirely agrees, so  far as it goes, with the views we have opened,   ' has the following remark in his last work, the  third volume of the Philosophy of the Human   ' Mind : " If a system of rational Logic should  ever be executed by a competent hand, this **  (viz. language as an instrument of thought)  '* will form the most important chapter."  Our doctrine is, that this will not merely form  the most important chapter, but that it wtU  be the only chapter strictly belonging to Jjo^ I  ^c ; and yet the theory we offer keeps deaf  of the extreme which betrayed Home Tooke,  who appears to consider reason as the result  of language. We pretend, then, to have inade  the discovery which Locke felt to be necessary,  and the nature of which Stewart more than i  conjectured j but oura is only " «?i Outline ; '*  and the system of rational Logic which the  Scotch metaphysician speaks of, yet remains to  be "executed by a competent hand:" — we     ON RHETORIC. [CHAP. III.   pretend but to have ascertained for it the  true foundation. — Something might be add-  ed on the importance which the subject de-  rives from the aspect of the times : for the  most careless observer cannot but remark,  how the rapid communication of knowledge  from mind to mind moulds and forms public  opinion ; and how the opinion of the many, ac-  quiring, day by day, a character and a weight  that never distinguished it before, threatens to  become the law to which not only individuals,  but governments, and eventually the common-  wealth of nations, must conform ; and hence we  might be led to urge that Philosophy cannot  be employed more opportunely, than in a new  examination of the instrument by which so  much has been, and so much more is likely  to he effected. The consideration is, how-  ever, too obvious not to have occurred to the  reader, and we therefore close our remarks.     CORRIGENDA ET ADDENDA.     At page 55, the assertione, that the words of a sen-  tence, " as parts of that sentence'''', and the sentences  of a discourse, " na parts of that discourse"", are not  by themselves significant, would perhaps sound a little  less paradoxical, if, instead of each of the phrases quo-  ted, the reader were to substitute " as parts of that  completed expression ".   At page 88, supply the other parenthetical mark  after " imderstanding" in line 4.   At page 196, line 6, the question is asked, whether  the juryman must go to Aristotle, and be taught to  compare the middle with the extremes ? The reader  will observe that the example is already farced into a  form, namely that of a syllogism in barbara, which a  juryman untaught by Aristotle would probably never  think of giving it, the other way of speaking being by  far the more obvious, viz. To kill a man maliciously  is murder ; A. B. killed a man maliciously ; therefore  A. B. is guilty of murder. Here, instead of the Aria-  totclian names major and minor, we prefer calling the  first proposition the datum, and the second, with re-  ference to the datum it is addressed to, the argument ;  and the truth of the argument having been proved by  testimony, we atfirm that the conclusion is as evident  as a conclusion can be, and that the Aristotelian  formula is a needless and puerile addition to a process  already complete — a proof of what is proved : — it is a  use of language for the purpose of reasoning which  does not identify with, but goes beyond, and childishly     252     CORRIGENDA ET ADDENDA.     refines upon that use of language in which the logic  of mankind at large consiets.   The doctrine of the whole work may receive some  light from the following way of stating it : — Man, in  common with other animals, derives immediately from  nature the power to express hie immediate, or, as they  are commonly called, his natural wants and feelings.  But he also possesses the power of inventing or learn-  ing a language which nature does not teach ; and it is  solely by the exertion of this power, which we call  reason, that he raises himself above the level of other  animals. By media such as artificial language consists  of, and only by such media, he acquires the knowledge  which distinguishes him from other creatures ; and  each advance being but the step to another, he is a  being indefinitely improveable. But if words are the  means of knowledge, it is an error to describe or con-  sider them in any other light ; and we accordingly  deem them not as, strictly speaking, the signs of  thought, but as the means by which we think, and set  others a thinking. This principle being admitted, ren-  ders unnecessary Locke's doctrine of ideas ; and Se-  MATOLOGY Stands opposed to, and takes the place of,  what the French call Iuealogy,   With respect to these addenda, should the reader  ask, whether they are to be esteemed a part of our  WORD, we answer in the affirmative. We imagined  our woED complete ;- — if, on fiirther consideration, we  had supposed so, we should not have added another  SYLLABLE : {^uT^Qh a ffvMMiiSavuv.)   G. WoedbUi Frlnlei, Angd Courl, SkJnnsi Street, Londoo.