III

ARS RHETORICA

For all a rhetorician’s rules
Teach nothing but to name his tools.

Samuel Butler, Hudibras 1.1.89–90

IT IS now time to give an account of the Art of Rhetoric as it was taught in the schools and expounded in the textbooks when rhetoric became established at Rome. It had grown up in the course of centuries and there were variations between the versions taught in the different schools. But the main outlines remained much the same, and many of the details were handed down unchanged from generation to generation. To most persons today this system is of little interest. It appears as an attempt by professional teachers to construct an obscure and difficult science out of the obvious. And so perhaps, in part at any rate, it is. Yet some of the best intellects of Rome did not think so. A study which engaged the attention of Cicero and Quintilian and which trained many Roman statesmen and men of letters cannot be entirely neglected, however repellent or frivolous it may seem to those brought up in different disciplines and systems of thought.

Complete systems of rhetoric are given in two works of the Republican period, Ad Herennium and Partitiones Oratoriae. In addition to these we have Cicero’s De Inventione, covering one division of the whole, parts of De Oratore, which though deliberately unsystematic, is based on the main divisions of traditional rhetoric, and Topica which deals with an obscure province on the borders of rhetoric and logic. From the Imperial period we have the massive work of Quintilian, which sums up with great thoroughness ancient rhetoric down to his day, and the lesser rhetoricians of the fourth and fifth centuries A.D. In the following outline the treatise Ad Herennium will be generally followed;it is the most convenient guide, for it is relatively simple and it best represents the traditional rhetoric of the schools, the sort of thing that was taught by the ordinary rhetorician in Rome. De Inventione is a fuller and more careful, or one might say, more pedantic version, with considerable variations, of the most important part of the system of Ad Herennium. Partitiones Oratoriae follows a different method devised in the Academy.¹ Whatever its merits, it was not generally followed, and it had little influence on later rhetoric.²

The material of rhetoric was traditionally divided according to the different types of oratory, the different functions of the orator and the different parts of a speech. The three types of oratory are forensic or judicial, deliberative and epideictic (genus iudiciale, genus deliberativum and genus demonstrativum), the first concerned with accusation and defence in the courts, the second with the recommendation of a course of action in deliberative assemblies, the third with praise and blame.³ In the Hellenistic rhetoric which Rome adopted the three types were given very unequal treatment; the rhetoricians had much to say about forensic oratory and little about deliberative, while epideictic was scarcely recognised as having an independent existence.

The functions of the orator (officia oratoris) or, as they were sometimes called, the parts of rhetoric,⁴ are firstly inventio, invention, or, as the textbooks defined it, ‘the devising of matter true or probable which will make a case appear more plausible’;⁵ secondly, dispositio, arrangement; thirdly, elocutio, style or presentation; fourthly, memoria, memory, and fifthly, actio or pronuntiatio, delivery. The last two stand apart as being matters more of nature than of art, and some authorities omitted one or both of them from the list.⁶

The parts of a speech (partes orationis) are the exordium or opening, the narratio or statement of facts, the divisio or partitio, that is, the statement of the point at issue and exposition of what the orator proposes to prove, the confirmatio or exposition of arguments, the confutatio or refutation of one’s opponent’s arguments, and finally the conclusio or peroration. This sixfold division is that given in De Inventione and Ad Herennium,⁷ but Cicero tells us that some divided into four or five or even seven parts,⁸ and Quintilian regards partitio as contained in the third part, which he calls probatio, proof, and thus is left with a total of five.⁹

Let us now see how the rhetoricians dealt with the different parts of the speech. Firstly, the exordium. The student was first informed that there were four types of case, honourable, base, doubtful and mean,¹⁰ and that the exordium must suit the type; then that there were two types of exordium, the principium which performed its function without any beating about the bush, and insinuatio, the indirect approach, to be used when the speaker was for various reasons in an unfavourable position.¹¹ The orator’s object in the exordium, and in particular in the first variety, was to make his listeners attentive, ready to learn and well disposed. Various precepts were given as to how this should be done;¹² for instance he was advised to secure attention by claiming that he was about to speak of important, new and unprecedented matters. At this point we have our first introduction to the doctrine of loci, the ‘places’ in which arguments can be discovered. The loci for making the audience well disposed are to be found in the speaker’s character, in that of his opponent, in that of the audience and in the actual facts of the case.¹³

The precepts for insinuatio were of a more practical character.¹⁴ When the case to be pleaded was discreditable, the audience was first to be reconciled to it by various means; the speaker would say that he shared his opponent’s dislike of whatever the latter had objected to, but that it bore no relation to himself. He would disclaim the intention of saying anything about his opponent, but would do so indirectly. He would aim not at openly damaging a popular figure but at undermining his position. If his opponent appeared to have made a strong impression, he could say that he was going to demolish the other’s strong points first, could express doubts as to where to begin, and so on. In cases where the audience was tired, he could begin with some amusing extemporary reference, say he would speak more briefly than usual, or adopt some other similar technique.

We now come to the narratio,¹⁵ which, like so much in rhetoric, was divided into three classes. Firstly there was the straightforward exposition of the facts of the case, the narratio in the usual sense of the term; then there were narrations which occurred incidentally in the course of the speech; thirdly there were those that had no connection with cases in the courts, this last category being divided, with typical pedantry, into those concerned with things and those concerned with persons.¹⁶ Detailed precepts were given only for the first type, which, it was laid down, should be brief, clear and plausible.¹⁷ The advice given under the last heading shows that the narratio was not necessarily an exact account of what actually happened.¹⁸ The author of Ad Herennium tells us that if the facts are true one should none the less observe the rules for plausibility, for the truth does not always carry conviction; if they are false, they need to be observed all the more.¹⁹ It was also suggested that when the events had to be explained away by the argument it would be best to dispense with a formal narratio and distribute the statement of facts throughout the speech.²⁰

The third part of the speech is the partitio or divisio; here the speaker would first state what points were agreed between himself and the other side and what was in dispute, then would give an orderly exposition of what he was going to say.²¹ There followed the confirmatio, which contained the most important part of the speech, the argument. This was treated with great elaboration by the rhetoricians; for it appealed to the logical interests of the Greeks, and lent itself to those elaborate analyses by which they made of rhetoric a serious intellectual discipline.

At this point²² we are introduced to the doctrine of the status or constitutio,²³ that is, the question from which the case arises, a doctrine particularly associated with Hermagoras. The status (there is no English term which represents the Latin satisfactorily) was conventionally defined as the ‘first conflict of the pleas arising from the answer to an accusation’, though, as Quintilian points out, its essence lay not in the ‘first conflict’ between one side and the other, but in the kind of question that arose from the conflict.²⁴ The author of Ad Herennium tells us that some considered there were four sorts of constitutio, whereas his teacher recognised only three.²⁵ He is referring to the four of Hermagoras, (coniectura), (finis), (qualitas) and (tralatio), the last of which was commonly rejected as an independent status. This classification gave rise to much argument and discussion among the more theoretically minded rhetoricians, as can be seen from Quintilian’s painstaking survey of the many different existing views.²⁶ But for practical purposes the threefold classification that the author of Ad Herennium adopted from his master was sufficient.²⁷ The three status could be conveniently distinguished by the three different questions which were involved, an sit, quid sit and quale sit. In the terminology of Ad Herennium they are the constitutio coniecturalis, where the dispute is one of fact, whether an action took place or not; the constitutio legitima, where the dispute arises from the interpretation of law; and the constitutio iuridicialis, where the act is admitted but the question is whether it was justified or not.

Having decided on the constitutio, we must, the rhetorician tells us, decide on the ratio, the line of defence; then we must consider the firmamentum, the answer which the accusation makes to this defence. Thus we arrive at the iudicatio, the , the question which arises from the defence and the counter argument of the accusation.²⁸ This piece of mystification appears to have originated with Hermagoras, who is rightly censured by Quintilian for his laborious and futile attempt at subtle definition.²⁹

Let us now see how the doctrine of inventio was applied to the different constitutiones and their subdivisions.³⁰ In the case of the constitutio coniecturalis, where the question is whether the action took place or not, six types of argument are distinguished.³¹ First there is that from probability. Under this head come possible motives and the man’s previous life. The two should indeed be combined: if the accuser says the motive of the accused was avarice, he will attempt to show that he has always been avaricious; if he cannot do this, he will find some other vice to attribute to him; if no such attribution is at all plausible, he will say one should consider deeds, not reputation. Similarly the defendant will demonstrate the integrity of his past life, if he can; if he is notorious for evil living, he will first say that false rumours have been put about by his opponent. If all else fails, he will say that he is not defending his morals before the censors, but answering a charge in a court of law. The second argument is that from comparison; the accuser, for example, will say that the deed could not have helped anyone except the defendant. Then comes that from circumstantial evidence, divided into six parts: place, time (e.g. at what hour was the deed committed?) duration (was there time for the deed to be completed?), opportunity, hope of fulfilment and hope of concealment. Fourthly there is the proof (argumentum) divided into past (e.g. where the accused was before the act, who his companions were, whether he said anything unusual), contemporary (e.g. whether he was seen at the time, whether any cries or other sounds were heard) and subsequent (e.g. whether anything was left behind to prove that the act was committed or to point to its author). Next comes consequence, consecutio, where one asks what are the usual consequences of crime and innocence. The accuser, for example, will say that his opponent showed all the signs of guilt, blushing, pallor, faltering, etc. If he did nothing of the sort the accuser will maintain that his demeanour was a proof not of innocence but of premeditation and brazenness. Finally there is approbatio, confirmation, which includes loci proprii and communes, topics peculiar to one side and those of general application.³² As regards the latter, commonplaces, there were considered to be two types, disquisitions on the heinousness of certain notorious sins and sinners, and certain general questions which can be argued both ways, such as the credibility of witnesses and the desirability of believing rumours.³³ The author of Ad Herennium is here only concerned with the second type, for which he duly provides the arguments on both sides. Thus in favour of rumours it may be said that there must be some foundation for the report; there was no reason why anyone should invent the story; other rumours may be false, but this one is true. On the other side we can show by examples that many rumours are false; we can say that the story has been invented by our enemies or by malicious and slanderous persons; we can produce some rumour discreditable to our opponent, if necessary inventing one, then say that we do not ourselves believe it, but that it shows how easy it is to spread abroad such stories.³⁴

We now come to the constitutio legitima, which, it will be remembered, is concerned with a dispute arising from the interpretation of law. Of this six varieties are distinguished: scriptum et sententia (the letter versus the meaning of the law), contradictory laws, ambiguity, definition, transference (e.g. where the competence of the court is disputed) and ratiocinatio, inference, where there is no law exactly covering the case.³⁵ In De Inventione the first three and the last in this list are treated separately from the constitutiones as controversiae quae in scripto versantur,³⁶ while the fifth, transference, is, in deference to Hermagoras, made a separate constitutio,³⁷ and the second constitutio is called definitiva.³⁸

Detailed suggestions were given for arguments on these different points. It will suffice to give by way of example those to be used for the intention as opposed to the letter of the law: The law-giver will be commended for only committing to writing as much as he thought necessary and leaving unwritten what he supposed could be understood; it will be said that it is the mark of a pettifogger to busy oneself with words and letters and neglect the intention; that what is written is impossible, or incompatible with law, custom, nature, or equity and right; that an opposite intention is either null or foolish or unjust or impossible or inconsistent with the intention of what has gone before or follows after, or disagrees with common law or with other common laws or with precedent. There should follow an enumeration of suitable examples to support the case.³⁹

The third constitutio, iuridicialis, was divided into two, absoluta and assumptiva.⁴⁰ The first is where we claim that what we have done is right absolutely (here it was thought necessary to introduce a classification of the different sources of right, ius),⁴¹ the second where the defence depends on extraneous considerations. Of this four varieties were distinguished; firstly there is concessio, where we admit the deed, but ask for pardon.⁴² This is divided into purgatio and deprecatio. In the case of purgatio we claim that the deed was not done intentionally, and put the blame on chance or inadvertence or necessity. Precepts were given under these three headings, but it was observed that the three were closely related and the same arguments more or less could be applied to each of them.⁴³ Deprecatio was where no such pleas could be made, and it was necessary to appeal for mercy on the ground of past services, good character and the like. Such a defence, it was laid down, was not suitable to the courts; it could however be used in the senate and other deliberative bodies.⁴⁴ So much for concessio.

Secondly there is tralatio criminis, when the deed is said to have been done as the result of someone else’s misdeeds;⁴⁵ thirdly the closely related remotio criminis, when some other person or thing is said to have been responsible.⁴⁶ Finally there is comparatio, where it is claimed that the course followed was better than its alternative.⁴⁷

Having set out the arguments to be used in each constitutio, of which we have given only a small selection here, the author of Ad Herennium proceeds to give some general precepts regarding argumentation.⁴⁸ We should not, he states, linger too long on the same topic, nor return again and again to the same point, nor leave an argument incomplete, nor move on to another in a clumsy fashion. The most complete and perfect form of argument is said to be a fivefold one, consisting of propositio, a statement of what we are about to prove; ratio, a brief exposition of our argument; confirmatio, a confirmation of this by a number of proofs; exornatio, embellishment; and complexio, summing up.⁴⁹

This section of Ad Herennium represents as much as the ordinary Latin teacher could assimilate of the general principles of argumentation as developed by the rhetoricians of Greece. For a general consideration of the topics of argument we turn to De Inventione.⁵⁰ Here we are told that arguments may be drawn either from the attributes of persons or from those of things. A list of eleven attributes of persons and four of things is given, in many cases with subdivisions. Thus the second class of attributes of things, those concerned with the actual performance of the deed, is divided into five—place, time, occasion, manner and facilities. Under the heading time we are told to consider not merely the past, the contemporary and the subsequent, but different types of past event, the fabulous past, recorded history and the recent past. Nor are we spared a definition of time,⁵¹ which takes us far from the forum to the tranquil atmosphere of the lecture room. We need not follow the youthful Cicero further into the various divisions and subdivisions of the different types of argument, for it may be supposed that such theoretical niceties were the part of the system heard with least patience by the average Roman student.

The general consideration of the nature of argumentation led naturally on to logical fallacies and faulty arguments generally. A classification of these formed the substance of what the rhetoricians had to say on the penultimate part of the speech, the reprehensio or refutation of one’s opponent’s arguments. Even the practical author of Ad Herennium deals with this section from the theoretical point of view, giving a lengthy exposition of the various types of faulty argument, with illustrations, curiously enough, from the Roman poets.⁵²

We have now reached the last part of the speech, the conclusio or peroration.⁵³ This the rhetoricians divided into three, enumeratio, recapitulation,⁵⁴ amplificatio and commiseratio. Amplificatio, a word generally used by rhetorical writers in a wider sense,⁵⁵ is the name given by Ad Herennium to what De Inventione calls indignatio, the arousing of indignation, and the ten topics of the former treatise are included in the fifteen of De Inventione. Similarly in the case of commiseratio, appeals for pity, Cicero is more elaborate than the author of Ad Herennium, and solemnly enumerates no fewer than sixteen different themes. Both authors however end with the advice not to linger too long on the appeal to pity, adding the apophthegm ‘Nothing dries quicker than a tear.’⁵⁶

So far we have been concerned only with the genus iudiciale, which, as has already been said, was treated with disproportionate elaboration. We now proceed to the genus deliberativum.⁵⁷ Deliberations were divided into two classes, according to whether the choice was limited to two alternatives or not; or into three, according to whether the desirability of the action discussed was absolute or relative or mixed.⁵⁸ The aim of the deliberative speaker is stated to be utility,⁵⁹ which is divided into two parts, that concerned with security and that concerned with honour. The former has two aspects, force and cunning,⁶⁰ the latter two also, the right and the praiseworthy.⁶¹ The right (rectum) includes the four virtues, prudentia, iustitia, fortitudo, modestia. In the definition of these virtues and in the recommendations for treatment under these four headings we can see signs of Stoic influence.⁶² But Stoic morality was not entirely suited to the forum, and we notice that precepts are given for speaking on the other side.⁶³ Virtue is not indeed to be openly depreciated, but it can be said that the facts of the case do not admit of the pursuit of the ideal, or that virtue is something other than what its advocates have claimed. What they call justice, for example, can be made out to be cowardice or perverse generosity, what they call courage to be the irrational daring of the gladiator.⁶⁴ The orator was advised to build up his speech on the same lines as in forensic speeches and to use the same methods of argumentation; and in the case of a conflict between safety and honour topics for argument were supplied to both sides.⁶⁵

The third type of oratory is epideictic, the genus demonstrativum, which was conventionally defined by its subject matter, praise and blame, a definition hardly compatible with its claim to be a separate genus.⁶⁶ The topics of praise, which also serve, in reverse, as the topics of blame, are divided into mind, body and estate, to borrow the Prayer Book phrase. The author of Ad Herennium, among other precepts, lays down that in recounting the life of the person concerned the orator should begin with externals—family and education—go on to bodily characteristics, and then return to externals, considering what virtues were displayed in connection with them. His full and even pedantic treatment of this type of speech, beginning with the proem—to be drawn either from the speaker’s character or from that of the person praised or from that of the audience or from the subject matter—shows that he had in mind a complete speech of praise or blame. Yet he tells us that the genus demonstrativum rarely occurs in real life, though occasions for praise and blame often occur in forensic or deliberative speeches.⁶⁷ Cicero, who in De Inventione gives this genus very meagre treatment, seems inclined in De Oratore to banish it altogether on the ground that it is unnecessary to treat every minor type of speech as if it needed special precepts peculiar to itself. In the end however he cannot bring himself to omit this section of rhetoric.⁶⁸

So far we have dealt solely with one part of the rhetorical doctrine, that which concerns inventio. The second part, dispositio, can be briefly despatched.⁶⁹ The parts of the speech, as we have seen, were dealt with under inventio, which had thus encroached on the province of dispositio, so that little was left to be said under the latter head. We are however told that there are two types of arrangement, that according to the rules (the rules, that is, for the parts of the speech given under inventio) and that where the orator uses his judgment to vary the order.⁷⁰ Once you have admitted that the arrangement of the speech can be varied according to circumstances and that the orator should use his judgment, there is little that can be given in the way of instruction. It was however generally agreed that as regards arrangement of arguments it was best to put the strongest first and last, and the less strong in the middle, but, as Quintilian observes, the position of the defence was different from that of the prosecution.⁷¹ It was indeed impossible to lay down general rules to cover all cases. Quintilian justly observes that no one case is exactly like another, and that the pleader must rely on his own judgment. This did not prevent him from devoting a whole book to dispositio.⁷²

On elocutio there is more to be said. When we come to this branch of rhetoric we are conscious of a change of emphasis. We are not here so much concerned with utilitarian ends or with logical subtleties as we are in the first branch, inventio. Inventio was drawn in part from the professional handbook, in part from philosophy; in elocutio a new source is discernible, that of literary criticism. The doctrine of style had been developed by those whose interest was less in the forging of weapons for victory in argument than in the analysis of existing works of oratory from the formal point of view. Such appears to have been the origin of two classifications which found a place in school rhetoric, that of the three styles (genera dicendi) and that of the four virtues of style (virtutes dicendi), the former possibly, the latter undoubtedly, of Peripatetic origin.⁷³ The three styles appear in Cicero as the genus grande, medium and tenue, whereas Ad Herennium uses the terms grave, mediocre and attenuatum.⁷⁴ They played a minor part in Roman rhetoric until Cicero in the Orator made adroit use of them to show that the best orators, such as himself, had all three at their command.⁷⁵ The four virtues of style appear in De Oratore much as Theophrastus enunciated them;⁷⁶ in Ad Herennium the first two, correctness and clarity, are regarded as subdivisions of elegantia, and appropriateness is replaced by compositio, euphony.⁷⁷ It thus appears that the four virtues had got somewhat confused in the school tradition, whereas Cicero restored them in their original simplicity and clarity.

The major part of what the rhetoricians had to say on elocutio was concerned with ornatus, and in particular with those departures from normal usage which served to embellish and enliven the orator’s language, and which provided the rhetoricians with such a fruitful field for classification and definition. How much laborious thought they devoted to this branch of their subject we can see from Quintilian’s lengthy discussion of the distinction between tropes, figures of speech and figures of thought.⁷⁸ On the whole they shed little light for all their labours, and wasted much effort on introducing confusing innovations of terminology and on discovering figures in modes of thought and speech which hardly deserved to be so called.⁷⁹ It is with a certain relief that we turn to the unpretentious treatment of the subject by the author of Ad Herennium, who knows only the division into exornationes verborum and exornationes sententiarum, and is content to give under these two heads a long list of technical terms followed by definitions and examples.⁸⁰ Unlike Quintilian, for example, he has nothing of the literary critic about him; his eye is on the oratorical effect of the various figures.⁸¹

It would be tedious and unprofitable to follow him through his list of figures,⁸² but it may be of interest to select a few which he regards as particularly effective and to reproduce some of his comments.

The figure repetitio, the repetition of a word at the beginning of a series of clauses, better known by its Greek name anaphora, is described as full of charm, weighty and pungent, and is recommended as an embellishment and a means of rendering speech effective.⁸³ The allied figure conduplicatio the doubling of a word or words, is said to have a great effect on the hearer, being compared to a wound inflicted more than once in the same place.⁸⁴ The collecting together of a number of points scattered throughout a speech, in particular of the various grounds for suspecting an accused person (frequentatio), is a vehement exornatio, almost essential in the constitutio coniecturalis, and useful in other types of case.⁸⁵ Permissio, a figure by which the orator claims to be handing himself over to someone else’s power—says in effect ‘Do what you like to me’—is recommended as particularly good for arousing pity.⁸⁶ Pity and indignation alike may be aroused by a lively description of the consequences of an action (descriptio).⁸⁷ This figure was distinguished from demonstratio , a vivid word picture of events, useful not only for commiseratio but also for amplificatio.⁸⁸ The same purposes were served by conformatio, more commonly known as prosopopoeia, the putting of a speech in the mouth of some character imagined as present or of some inanimate object or abstraction endowed with life, making the commonwealth speak, as Cicero puts it, and calling up the dead from the underworld.⁸⁹ This figure Cicero regarded as exceptionally effective.⁹⁰

Two figures which the rhetoricians coupled together were simile (similitudo) and example (exemplum).⁹¹ The lengthy treatment of the simile in Ad Herennium shows that its use was carefully taught in the rhetorical schools. The would-be orator is told that the discovery of similes will be easy if he conjures up before his eyes everything, animate or inanimate, dumb or possessed of speech, wild or tame, belonging to earth, sky or water, manufactured, accidental or natural, common or uncommon; from this wide field he can hunt out some simile which will serve one of the four purposes of the figure, embellishment, proof, clarity or vividness.⁹² The exemplum, which serves the same four purposes as the simile, is the adducing of an action or saying ascribed to a particular person.⁹³ The two figures are described by Cicero as particularly moving.⁹⁴

Some figures were recommended only with qualification. Exclamatio (apostrophe), the addressing of a person, city, place, etc., was to be used sparingly, and only when the importance of the matter justified it.⁹⁵ Very sparing use was advised in the case of similiter cadens , the use of a series of words in the same case; of similiter desinens , the use of words with similar terminations;⁹⁶ and of annominatio, word play.⁹⁷ Such figures are witty and amusing rather than dignified and beautiful, and give the impression of having been laboriously thought up beforehand.⁹⁸ The sententia, the brief statement of a general truth, of which Ad Herennium distinguishes two types, the short and the longer, is to be used sparingly, ut rei actores, non vivendi praeceptores videamur.⁹⁹

Finally we come to the last two parts of the rhetorical system, actio and memoria. There is no space here to describe the various aids to memory prescribed by the rhetoricians,¹⁰⁰ but something must be said of delivery, for this was a side of the art to which the ancients attached great importance.¹⁰¹ If the author of Ad Herennium is to be believed, none of his predecessors had treated the subject in a systematic manner, on the ground that its nature was such that it could hardly be written about.¹⁰² It may therefore be that he is not here representative, and that some students of rhetoric had the good fortune to escape the pedantic method of treating the subject which he followed. There is little trace of the textbook in Cicero’s treatment in De Oratore, nor indeed in Quintilian, in spite of his elaborate descriptions of the uses of the various facial expressions, bodily movements, gestures of hands and arms, and so on. His long and entertaining chapter on delivery¹⁰³ should be read by all who want to understand ancient oratory. But as our purpose here is to summarise what was laid down in the textbooks, we will continue to follow Ad Herennium as our guide.

The author divides the subject of actio into two parts, voice production and bodily action. As regards voice, the three qualities required were strength, durability and flexibility. The first was mainly a matter of nature, but the last two could be cultivated by exercise.¹⁰⁴ When we come to flexibility we are met with the inevitable tripartite division, into sermo, the conversational manner, contentio, the argumentative, and amplificatio, the emotional. The first has four subdivisions, the dignified, the instructive, the narrative and the humorous; the second has two, the rapid and the abrupt; the third also has two, the pathetic and the indignant.¹⁰⁵

We will pass over the precepts for voice training, and proceed to the precise and curious directions on bodily action, which includes not only movement of the limbs, but also facial expression.¹⁰⁶ In using the colloquial manner, dignified or narrative, the orator will stand on the soles of his feet and adopt a gentle movement of his right hand, while his facial expression will be in accord with the mood of his language; in the instructive-colloquial he will bend forward a little; in the humorous he will express the mood by his face, not by gestures. In the continuous argumentative style the arm is moved rapidly, the facial expression is mobile, the look piercing; in the abrupt, the arm will be thrust forward rapidly, the orator will walk about, occasionally stamp his right foot,¹⁰⁷ and put on a piercing and intent look. When he is arousing indignation he will proceed as in the case of continuous argument, except that the gestures will be slower; in the pathetic style it will be appropriate to strike the thigh and beat the head,¹⁰⁸ and in many cases to use gestures subdued and firm and an expression sad and disquieted.

Such was the Art of Rhetoric. It bears all the marks of its Greek origin. It was the creation of the Greek intellect, with its intellectual subtlety and its love of logic and fine distinctions, but of the Greek intellect in its decline, no longer adventurous and creative, but confined within professional and pedagogic bounds. It was open to criticism, and it was sometimes criticised, for its pedantry and remoteness from life. But on the whole the Romans took to it with surprising readiness. From its first introduction to the end of Roman civilisation it was part of the intellectual background of the educated Roman.