always live, meaning his works, the cause is tranf. ferred to the effect. All words muft at firft have, had one original and primary fignification, which, ftrictly speaking, may be called their proper fenfe. But through length of time, words lofe their original fignification, and affume a new one, which then becomes their proper fenfe. For example the word knave anciently fignified a fervant.

127. Sometimes two or more words have the fame fignification, and may therefore be used fynonymously. Of this kind are the words pity and compaffion. There are other words of fo near an affinity to each other, that they are commonly thought fynonymous. Such are mercy and pity; though mercy in its strict sense is exercifed towards an offender, and pity respects one in diftrefs. This peculiar force and diftinction of words is carefully to be attended to. Though words, when taken in their proper fignification, generally convey the plaineft and clearest sense; yet some are more forcible, fonorous, or beautiful, than others. And by these confiderations we must often be determined in our choice of them. So whether we say, be got, or be obtained, the victory, the fenfe is the fame; but the latter is more full and fonorous.

128. As to tropes, though generally speaking they are not to be chofen where perfpicuity is only defigned, and proper words may be found; yet through the penury of all languages, the ufe of them is often neceffary. And fome of them, ef pecially metaphors, which are taken from the fimilitude of things, may, when cuftom has rendered them familiar, be confidered as proper words, and used in their ftead. Thus, whether we fay, I fee your meaning, or, I understand your meaning, the fenfe is equally clear, though the latter expreffion is proper, and the former metaphorical.

129. ii. But perfpicuity arifes not only from a choice of fingle words, but likewife from the confruction of them in fentences. For the meaning of all the words in a sentence, confidered by themselves may be very plain and evident; and yet, by a diforderly placing them, or confufion of the parts, the fenfe of the whole may be obfcured. It is certain that the most natural order is the plaineft, when both the words and parts of a fentence are fo difpofed as beft agrees with their mutual relation and dependence upon each other,

130. AMBIGUITY of expreffion is one caufe of obfcurity. This fometimes arifes from the different fenfes in which a word is capable of being taken. In all ferious difcourfes ambiguities ought carefully to be avoided. But obfcurity more frequently arifes from the ambiguous construction of words, which renders it difficult to determine in what fenfe they are to be taken. Quintilian gives us this example: "A certain man ordered in his will, that his heir fhould erect for him a ftatue holding a fpear made of gold." A queftion arofe here, of great confequence to the heir from the ambiguity of the expreffion, whether the words made of gold were to be applied to the flatue or the Spear?

131. OBSCURITY is also occafioned either by too short and concise a manner of speaking, or by fentences too long and prolix: and parentheses, when either too long or too frequent, have the fame effect.

SECT. II. Of COMPOSITION.

132. COMPOSITION, in the fenfe here used, gives rules for the ftructure of sentences, with the feveral members, words, and fyllables, of which they confift, in fuch a manner as may beft con. tribute to the force, beauty, and evidence of the whole.

133. Compofition confifts of four parts, which rhetoricians call period, order, jun&ure, and number. The first treats of the ftructure of sentences; the fecond of the parts of fentences, which are words and members; and the two last of the parts of words, which are letters and fyllables. For all articulate founds, and even the moft minute parts of language, come under the cognizance of oratory.

134. I. Of PERIOD. In every fentence or propofition, fomething is faid of fomething. That of which fomething is faid, logicians call the fubje&, and that which is faid of it, the predicate: but in grammatical terms, the former is a noun fubftantive, and the latter a finite verb, denoting affirmation, and fome ftate of being, acting, or fuf. fering.

135. A fimple fentence confifts of one fuch noun and verb, with whatever elle is joined to either or both of them. And a compound fentence contains two or more of them; and may be divided into fo many diftinct propofitions, as there are fuch nouns and verbs, either expressed or understood.

136. Some fentences confift, either wholly or in part, of fuch members as contain in them two or more compound ones, which may therefore, for diftinction's fake, be called decompound mem bers. In the proper distinction and feparation of the members in fuch complex fentences, the art of pointing chiefly confifts. For the principal ufe of a comma is to divide the fimple members, a femicolon the compound ones, a colon fuch as are decompounded, and a period the whole from the following fentence. See PUNCTUATION.

137. Sentences, with refpect to their form or compofition, are diftinguished into two forts, called by Cicero tracta, ftraight or direct; and contorta, bent or winding. By the former are meant those whofe members follow each other in a direct order, without any inflection; and by the latter, thofe which strictly speaking are called pe riods. For wig in Greek fignifies a circuit or circle: And fo the Latins call it circuitus and ambitus: By which both mean a sentence confifting of correfpondent parts, fó framed, that the voice in pronouncing them may have a proper elevation and cadency, and diftinguish them by its inflection; and as the latter part returns back, and unites the former, the period, like a circle, furrounds and inclofes the whole fenfe. This elevation of the voice, in the former part of the period, is by the Greeks called #poraris and by the Latins propofitio; and the depreffion of it in the latter part, by the one arodoris, and by the other redditio.

138. As fimple fentences have not these corref pondent parts which require any inflection of the voice, nor a circular form, by reason of their bre vity, they are not properly periods, in the ftric

fenf

fenfe of the word: though, in common fpeech, the words fentence and period are used as equivalent terms. As to thofe compound fentences whofe members follow each other in a direct order without any inflection, there is little art required in their compofition. Since fuch fentences, therefore, may be limited at pleasure, it seems more convenient, both for the speaker and hearers, to confine them to a moderate length.

139. But the principal art relating to this part of compofition lies in the frame and structure of fuch compound fentences as are properly called periods. In the formation of these, two things are chiefly to be regarded; their length and cadency. As the length ought to be fuited to the breath of the speaker, the ancient rhetoricians scarce admit of more than four colons; by which we may understand compound members of a moderate fize, which will be generally found a fuitable and proportionate length. As to the cadency, Cicero has obferved, that the ears judge what is full and what is deficient; and direct us to fill up our periods, that nothing be wanting of what they expect. When the voice is raifed at the beginning of a fentence, they are in fufpenfe till it be finished; and are pleated with a full and juft cadency, but are fenfible of any defect, and difpleafed with redundancy. Therefore care must be taken that periods be neither deficient, and as it were maimed, fo as to defraud the ears of what feemed to be promised them; nor, on the other hand, of fend them by too long and immoderate excurfions. This rife and cadency of the voice in pronunciation, depend on the nature and fituation of the members.

140. By the word members are to be underftood fuch as are uncompounded. In a period of two members the turn of the voice begins with the latter member. If a sentence confift of three members, the inflection is best made at the end of the 2d member: for if it begin immediately after the first, the voice will be either apt to fink too low, and not to be heard, before it reach the end; or elfe be precipitated, to prevent it. But a period of four members is reckoned the most complete and perfect, where the inflection begins at the middle, that is with the 3d member. Nor is it the fame cafe here, as if, in a fentence of three members, the cadency be made at the 2d; for in proportion to the time of raising the voice may the space be allowed for its finking. But where the sense does not fuit with this divifion at the entrance upon the 3d member, it should be made at the 4th.

4

fpeaking. And in this view they are confidered and made ufe of by the orator.

142. Examples of all these might be quoted from Cicero and other ancient orators; but it is unnecessary to swell this article with long quotations. The English rder can hardly take up a newspaper, during the fitting of the Imperial Parliament, in which he will not find numerous examples of all the beauties of modern oratory, and the learned can easily confult the orators of ancient Greece and Rome.

143. II. Of ORDER. BY ORDER rhetoricians mean the placing each word and member of a fentence in fuch a manner as will most contribute to the force, beauty, or evidence of the whole.

144. Order is of two kinds, natural and artificial; and each of these may be confidered with refpect to the parts, either of fimple or compound fentences.

145. i. As to fimple fentences, we call that erder NATURAL, when all the words in a fentence are fo placed, as they are connected with or follow each other in a grammatical conftruction.

146. ii. ARTIFICIAL ORDER, as it respects fimple fentences, has little or no regard to the natural conftruction of words; but disposes them in fuch a manner as will be moft agreeable to the ear, and beft answer the defign of the speaker.

147. As to compound fentences, or fuch as confift of two or more members, either fimple or compounded; what relates to the words in each member feparately is the fame as in fimple fentences. But with regard to the difpofition of the several members, that is called the natural order which fo places them as they mutually depend on each other. When this order is inverted, it is styled artificial.

148. It is unneceffary to enlarge farther, or to adduce examples upon this fubject, as it is fully discussed, and the beauties as well as defects of the analagous and tranfpofitive languages pointed out under the article LANGUAGE; Sect. IV-VI.

149. But there are some other confiderations relating to order, which, being taken from the nature of things, equally fuit all languages. So, in amplifying, there fhould be a conftant gradation from a leis to a greater; as when Cicero fays, Ambition creates hatred, fhyness, difcords, feditions, and wars. On the contrary, in extenuating, we should defcend from a greater to a lefs; as if, fpeaking of the ancient laws of Rome, one should fay, They were fo far from fuffering a Roman citizen to be put to death, that they would not allow him to be whipt, or even to be bound. In conftituting any whole, we put the parts firft; as, Invention, difpofition, elocution, and pronunciation, make up the art of oratory. But in feparating any whole, the parts follow: as, The art of oratory may be divided into thefe four parts; invention, difpofition, elocution, and pronunciation. In every enumeration care must be taken not to mix the whole with the parts; but if it be mentioned at all, it must either be put firft or laft. So it would be wrong to say, He was a man of the greatest prudence, virtue, juftice, and modefty: for the word virtue here contains in it the other three, and therefore (hould not be inferted among them. See LANGUAGE, Se&. VII.

Mmm 2

#50. 111

150. III. JUNCTURE and NUMBER relate to letters and fyllables; the former treating of their connection, and the latter of their quantity.

151. JUNCTURE refpects the quality of fyllables. A due attention is to be paid to the nature of the vowels, confonants, and fyllables in the connection of words, with gard to the found. When a word ends with a vowel, and the next begins either with à different vowel, or the fame repeated, it usually renders the pronunciation hollow and unpleasant. Those which have the weakeft and smallest found follow beft; because they occasion the leaft alteration of the organ in forming the two founds. If a word end with a vowel, the next ought to begin with a consonant, or such a vowel whofe found may agree well with the former. But if a word conclude with a confonant, either a vowel should follow, or such a confonant whofe proper pronunciation will suit with it. And the fame fyllable ought not to be repeated at the end of one word, and the beginning of the next. This laft, however, rarely happens on our language, which abounds with confo

nants.

152. IV. NUMBER refpects the quantity of fyllables. In the Greek and Roman languages every fyllable has its distinct quantity; and is either long, fhort, or common two or more of which joined together in a certain order make a foot, and a determinate number of thefe in a different order conftitute their feveral forts of verfe. Though their profe was not fo confined with regard to the feet, either as to the kind or place of them, as their metrical compofitions, yet it had a fort of measure, more especially in the rife and cadeney of their periods. This they called rhetorical sumber. But their rules are not applicable to our Janguage, which has not that accurate diftinc tion of quantity in its syllables.

153. A great number of monofyllables do not found well together. For as there ought to be a greater distance in the pronunciation between one word and another, than between the fyllables of the fame word, fuch pauses, though fhort, yet, when too frequent, make the found rough, and thus fpoil its harmony. This is the more neceffary to be attended to, because the English language abounds with monofyllables. On the contrary, a continuation of many long words makes a fentence inove too flow and heavily. And, therefore, fuch periods generally run beft, which have a proper mixture of words of differ ent lengths. Befides, as every word has its accent, which with us ftands for quantity, a num ber, either of monofyllables or long words, coming together, abates the harmony, as it leffens the variety.

154. Several words of the fame ending do not found well together, especially where the accent falls upon the fame syllable in each of them. For this creates too great a jingle by the fimilitude of found; and difpleafes from an appearance of affectation. Of this kind is the following fentence; Nothing is more welcome, delightfome, or wholefome, than reft to a wearied man. In fuch expreffions, if the order of the words cannot well be altered, fome other word fhould be fubftituted in the room of one of them at lealt, to diverfify the found,

Thus the found might be varied by faying, Nothing is more welcome, pleasant, or wholefome.

155. But if a sentence end with a monofyllable, it is apt to hurt the cadency, and disappoint the ear; whereas words of a moderate length carry a greater force with them by the fulness of their found, and afford the ear what it expected. There is one fort of monofyllables efpecially, which never found well at the conclufion of a period, viz. the figns of cafes. Thus we fay, Avarice is a crime which wife men are too often guilty of. But the cadency would be more agreeable thus: Avarice is a crime, of which wife men are too often guilty. Nor are very long words well fuited either to the beginning or conclufion of a period; for they retard the pronunciation at first, and fall too heavy at the end.

SECT. III. Of DIGNITY.

156. DIGNITY confifts in the right use of tropes and figures. It is not fufficient for an orator to exprefs himself with propriety and clearness, or in fmooth and harmonious periods; but his language fhould be fuited to the nature and importance of the subject. And therefore, as elegance gives rules for the first of thefe, and compofition for the ad, so does dignity for the laft. It is evident, that different fubjects require a different ftyle and manner of expreffion; for, as Quintilian fays, "What is magnificent in one difcourfe would be turgid in another." This variety in the manner of expreffion arifes in a great measure from tropes and figures, which not only enliven and beautify a difcourse, but give it likewife force and grandeur.

157. TROPES and FIGURES are thus diftinguifhed: Tropes moftly affect fingle words, but figures whole fentences. A trope conveys two ideas to the mind by means of one word; but a figure throws the fentence into a different form from the ufual manner of expreffion. Tropes are chiefly defigned to reprefent our thoughts, but figures our paffions.

I. Of TROPES.

158. A trope is a figure of words, and has been ufually defined to be the change of a word from its proper fignification to fome other with advantage, either as to beauty or ftrength. But this definition is not strictly juft; for, in every trope a reference is had to two things, which occafions two ideas; one of the thing expreffed, and another of that to which it has a refpect. For all tropes are taken, either from things internally or externally related; or from fome fimilitude between them; or from a contrariety. The firft of thefe is called fynecdoche, the ad metonymy, the 3d métaphor, and the laft irony. The reafons which occafioned the introduction of tropes, as Quintilian observes, are three; necefity, emphafis, and beauty.

159. i. Tropes were first introduced from receffity, deriving their origin from the barrenness of language; because no language contains a fufficient number of proper words to exprefs all the differ ent conceptions of our minds: but the principal caufe of their introduction feems to be that extenfive influence which imagination poffeffes over every kind of fpeech. The mind confiders the fame thing various ways; views it in different

lights;

lights; and compares it with other things. Hence it is furnished with an almost infinite number of ideas, which cannot all be expreffed by proper words, fince new ideas occur daily. And were this poffible, yet would it be impracticable; becaufe the multitude of words must be fo vaftly great, that the memory could not retain them, nor recal them as occafion required. Tropes have redreffed both thefe inconveniencies. Thus, where a word is wanting to express any particular thing, it is clearly reprefented by the name of fome other thing refembling it; the caufe is fignified by the effect, the fubject by the adjunct, or the contrary; and the whole is often understood by a part, or a part by the whole. Thus, by the ufe of tropes, the mind is helped to conceive of fomething not expreffed, from that which is expreffed.

160. ii. A fecond ufe of tropes is Emphafis. Tropes often exprefs things with greater force than cati be done by - bper words. Sometimes a lively trope conveys a fuller and more juft idea of a thing than a large periphrafis. Thus, when Virgil calls the Scipios to thunderbolts of war, he gives a more lively image of the rapid force and speedy fuccefs of their arms than could have been conveyed by a long defcription in plain words. And in many cafes the tropical use of words is fo emphatical, that in this respect it may be justly efteemed the moft proper. So, incenfed with ciger, inflamed with defire, fallen into an error, are all metaphorical expreffions, yet perhaps no proper words can be made ufe of, which will convey a more lively image of the thing we defign to reprefent by them.

161. iii. Beauty and ornament have been another caufe of the ufe of tropes. Some fubjects require a more florid and elegant drefs than others. And it is the bufinefs of an orator to entertain his hearers while he inftructs them. Cicero has obferved, that "as garments were firft invented from neceflity, to fecure us from the injuries of the weather, but improved afterwards for ornament and diftinétion; fo the poverty of language firft introduced tropes, which were afterwards increased for delight."

162. As to the choice of tropes, they ought neither to be obfcure nor too far-fetched, left they should become unintelligible. It is neceffary, in a good trope, not only that there be a near affinity between the two ideas, but likewise that this affinity be obvious. A trope fhould likewife bear a due proportion to the thing it is defigned to reprefent, fo as neither to heighten nor diminish the juft idea of it. Poets indeed are allowed a greater liberty in this refpect; but an orator fhould be modeft in his expreffions.

163. As a moderate ufe of tropes, juftly applied, beautifies and enlivens a difcourfe, fo an excess of them caufes obfcurity, by running it into abftrufe allegories and riddles. Tropes are not the ordinary drefs of our thoughts, but a foreign babit; and therefore he who fills his difcourfe with a continued feries of them, acts like one who appears in public in a ftrange drefs. Care fhould alfo be taken not to transfer tropes, from one language into another. For as they are often taken from the manners, cuftoms, and occurrences of particular nations; fo they may be very plain and obyious to thofe among whom they firft took

their rife, but altogether unintelligible to others who are unacquainted with the reason of them. 164. TROPES are divided into primary and fecondary.

165. I. PRIMARY TROPES are fubdivided into four fpecies; 1. Metaphor: 2. Metonymy: 3. Synec doche: 4. Irony.

166. I. A METAPHOR is ufually defined, A trope, which changes awords from their proper fignification to another different from it, by reafon of fome fimilitude between them. But it is certain that a word, when ufed metaphorically, does not alter its fignification, but retains its proper fenfe. Cicero calls a metaphor a fimilitude reduced to a fingle sword. It is a fimilitude, when I say of a man, he has acted like a lion; and a metaphor, when I fay, he is a lion.

167. The ufe of metaphors is as extenfive as univerfal nature. For there are fcarce any two things which have not fome fimilitude between them. However, they may all be reduced to four kinds. 1. Thofe taken from fimilitudes between animate beings. 2. Those borrowed from fimilitudes between inanimate beings. 3. Those taken from fimilitudes between inanimate things and animals: and, 4. Those which are efteemed the fineft and strongest, which ascribe life and a&tion to inanimate things.

168. However, care must be taken not to venture upon too bold and daring metaphors. Poets indeed claim greater liberty in this refpect. But the orator fhould reafon coolly, though ftrongly and forcibly; and not by theatrical reprefentations fo transport the mind as to take it off from reflection. Yet metaphors ought not to fink below the dignity of what they are defigned to exprefs; but the idea they convey fhould at least be equal to the proper word in the place of which they are fubfituted. All forced and harsh metaphors fhould be avoided, nor fhould they be too numerous; in a word, they ought not to be used, but either where a proper word is wanting, or where they are more fignificant or beautiful than the proper word. See METAPHOR, $2.

169. II. METONYMY, as defined by Quintilian, is the putting one word for another. But Voffius defcribes it more fully, when he calls it," A trope which changes the name of things that are naturally united, but in such a manner as that one is not the effence of the other." A word ufed as a metonymy changes its fenfe, and denotes fomething different from its proper fignification. Thus, when Mars is put for war, and Ceres for corn, they lofe their perfonal fenfe, and ftand for the effects of which thofe deities were faid to be the cause.

170. A metonymy is not fo extenfive as a metaphor, nor altogether fo neceffary; because nothing is faid by a metonymy, which cannot be expreffed in proper words; whereas metaphors are often used for want of proper words to exprefs fome ideas. However, metonymies enrich a difcourfe with an agreeable variety, and give both force and beauty to an expreffion. And fome metonymies, in common difcourfe, are more frequently made ufe of than the proper words in whofe room they are put. So, it is more ufual to fay, This is fuch a perfon's hand, or I know his hand, than his

quriting,