between good speaking and bad, or between that which
gave more pleasure and that which gave less, between
grave or solemn, and commonplace or colloquial, consisted
in something additional superimposed upon the canvas of
ordinary speech like an embroidery by a skilful orator.
These considerations led the Græco-Roman rhetoricians
to adopt the practice, like the Indians, who arrived at
the distinction independently, to distinguish the bare
(n) or purely grammatical form from another form
containing an addition which they called ornament,
Kóσμos ornatum est (Quintilian will serve, as typical of
all the rest) quod perspicuo ac probabili plus est.1

The notion of ornament as something added on from
outside forms the basis of the theory which Aristotle, the
philosopher of Rhetoric, gave of the queen of ornaments,
Metaphor. According to him the high pleasure aroused
by metaphor arises from the collocation of different terms
and the discovery of relations between species and genera,
producing "learning and knowledge by means of the
genus” (μάθησιν καὶ γνῶσιν διὰ τοῦ γένους), and that
easy learning which is the greatest of human pleasures,2
which amounts to saying that metaphor adds to the
concept under consideration a group of minor incidental
cognitions, as a kind of diversion and relief and pleasant
instruction for the mind.

Ornaments were divided and subdivided in a number Classes of of different ways. Aristotle (and previously Isocrates, ornament. rather differently) classified the ornaments which diversify bare or nude form, under the heads of dialect forms, substitutions and epithets, prolongations, truncations and abbreviations of words, and other departures from common usage, and, finally, rhythm and harmony. Substitutions were of four classes: species for genus; genus for species; species for species; and proportionate.3 After Aristotle, elocution was especially studied by Theophrastus and Demetrius Phalereus: Phalereus; these rhetoricians and their followers further solidified the classification of ornament 2 Rhet. iii. ch. 10. 3 Poet. chs. 19-22; cf. Rhet. iii. cc. 2, 10.

1 Quintil. Inst. orat. viii. ch. 3.

The concept of the Fitting.

by distinguishing tropes from figures (σxnμara) and dividing figures into figures of speech (σχήματα τῆς λέξεως) and of thought (TŶs diavolas), figures of speech into grammatical and rhetorical, and figures of thought into pathetic and ethic. Substitutions were divided into fourteen principal forms,-metaphor, synecdoche, metonymy, antonomasia, onomatopeia, catachresis, metalepsis, epithet, allegory, enigma, irony, periphrase, hyperbaton and hyperbole; each divided into subspecies and contrasted with its relative vice. Figures of speech amounted to a score or so (repetition, anaphora, antistrophe, climax, asyndeton, assonance, etc.); figures of thought to about the same number (interrogation, prosopopoeia, ætiopœia, hypotyposis, commotion, simulation, exclamation, apostrophe, aposiopesis, etc.). If these divisions have any value as aids to memory in relation to particular literary forms, considered rationally they are simply capricious, as is evidenced by the fact that many classes of the ornate appear now under the heading of tropes, now of figures; sometimes under figures of speech, then as those of thought; no reason for the alteration is given except the arbitrary caprice of an individual rhetorician which so decrees and disposes. And since one function which may be fulfilled by the rhetorical categories is to point out the divergence between two ways of expressing the same thing, one of which is arbitrarily selected as "proper," 1 it is easy to see why the ancients defined metaphor as "verbi vel sermonis a propria significatione in aliam cum virtute mutatio," and figure as conformatio quaedam orationis remota a communi et primum se offerenti ratione." 2

So far as we know, antiquity raised no revolt against the theory of ornament or of double form. We do sometimes hear Cicero, Quintilian, Seneca and others saying, Ipsae res verba rapiunt, Pectus est quod disertos facit et vis mentis, Rem tene, verba sequentur, Curam verborum rerum volo esse sollicitudinem, or Nulla est verborum nisi rei cohaerentium virtus. But these maxims did not bear the weighty meanings which we moderns might attach to 1 See above, pp. 68-69. 2 Quintilian, Inst. orat. viii, ch. 6; ix. ch. I.

them; they were perhaps in contradiction with the theory
of ornament, but as the contradiction was unheeded, it
was ineffective: they were the protests of common sense,
powerless to combat the fallacies of school doctrine.
Moreover, the latter was fitted with a safety-valve, a sage
contrivance to disguise its inherent absurdity. If the
ornate consisted of a plus, in what degree should it be
used? if it gave pleasure, must we not conclude that the
more it were used, the greater the pleasure derived?
would its extravagant use be attended by extravagant
pleasure? Herein was peril: instinctively the rhetori-
cians hastened to the defence, snatching up the first
weapon that came to hand, namely, the fitting (péπov).
Ornament must be used carefully; neither too much nor
too little; in medio virtus; as much as is fitting (ảλλà
πρéπον). Aristotle recommends a style seasoned with
“a certain dose” (δεῖ ἄρα κεκρᾶσθαί πως τούτοις); for
ornament should be a condiment, not a food (dvoμa,
où edeσμa).1 The fitting was a concept quite inconsistent
with that of ornament; it was a rival, and enemy,
destined to destroy it. Fitting to what? to expression,
of course; but that which is fitting to expression cannot
be called an ornament, an external addition; it coincides
with expression itself. But the rhetoricians contented
themselves with maintaining peaceful relations between
the ornate and the fitting, without troubling to mediate
them through a third concept. The pseudo-Longinus
alone, in answer to an observation of his predecessor
Caecilius that more than two or three metaphors must not
be used in the same place, remarked that a larger number
ought to be used where passion (τà πáðŋ) rushes headlong
like a torrent, carrying with it as necessaries (s
ȧvaykaîov) a multitude of such substitutions.2

ornament in

Preserved in the compilations of later antiquity (such The theory of as the works of Donatus and Priscian and the celebrated the Middle allegorical tract of Marcianus Capella), and in the com- Ages and pendia of Bede, Rhabanus Maurus and others, the theory

1 Aristotle, Rhet. iii. ch. 2; Poet. ch. 22.
2 De sublimitate (in Rhet. graeci, ed. Spengel, vol. i. § 32).

Renaissance.

of ornament passed to the Middle Ages. Throughout this period Rhetoric, Grammar and Logic continued to form the trivium of the schools. The theory was to some extent favoured in mediæval times by the fact that writers and scholars made use of a dead language; this helped to reinforce the idea that beautiful form was not a spontaneous thing but consisted in an addition or embroidery. Under the Renaissance the theory continued to flourish and was revived by study of the best classical sources; to the works of Cicero were added the Institutiones of Quintilian and the Rhetoric of Aristotle, with the host of minor Latin and Greek rhetoricians, amongst whom was Hermogenes with his celebrated Ideas, brought into fashion by Giulio Camillo.1

Even those writers who dared to criticize the organism of ancient Rhetoric left the theory of ornament unassailed. Vives lamented over the exaggerated subtlety of the Greeks" which had multiplied distinctions to infinity in this matter without diffusing light, but he never took up a definite stand against the theory of ornament. Patrizzi was dissatisfied with the insufficient definition of ornament given by the ancients; but he asserted the existence of ornaments and metaphors as well as seven different modes of "conjoined speech,”-narrative, proof, amplification, diminution, ornament with its contrary, elevation and depression. The school of Ramus continued to entrust Rhetoric with the "embellishment" of thought. Owing to the vast extension and intensification of life and literature in the sixteenth century, it would be easy to quote phrases, as we have done from ancient authors, asserting the strict dependence of speech upon the things it wishes to express, and lively attacks on pedants and pedantic forms and rules for beautiful speech. But what would be the use? The theory of ornament was always in the background, tacitly admitted as

3

1 Giulio Camillo Delminio, Discorso sopra le Idee di Ermogene (in Opere, Venice, 1560); and trans. of Hermogenes (Udine, 1594). 2 De causis corruptarum artium, loc. cit.

3 Della rhetorica, dial. 6.

indisputable by all. Juan de Valdés, for instance, makes
the following confession of stylistic faith: "Escribo como
hablo; solamente tengo cuidado de usar de vocablos que
sinifiquen bien lo que quiero decir, y dígolo cuanto más
llanamente me es posible, porqué, á mi parecer, en ninguna
lengua está bien la afectación." But Valdés also says that
beautiful language consists en que digais lo que quereis
con las menos palabras que pudiéredes, de tal manera
que... no se pueda quitar ninguna sin ofender á la
sentencia, ó al encarescimiento, ó á la elegancia." 1 Here
it seems that amplification and elegance are conceived as
extraneous to the meaning or content.-A gleam of truth
is visible in Montaigne, who, confronted by the laboured
categories into which rhetoricians divide ornament,
observes: "Oyez dire Métonymie, Métaphore, Allégorie et
aultres tels noms de la Grammaire; semble il pas qu'on
signifie quelque forme de langage rare et pellegrin? Ce sont
tiltres qui touchent le babil de vostre chambrière." 2 That is
to say, they are anything but language remote from the
primum se offerens ratio.

the seventeenth

The impossibility of upholding the theory of ornament Reductio ad was first noticed during the decadence of Italian literature absurdum in in the seventeenth century, when literary production century. became but a play of empty forms, and the "convenient,” long violated in practice, was abandoned and forgotten even in theory, and came to be looked on as a limit arbitrarily imposed on the fundamental principle of ornamentation. The opponents of that style loaded with conceits which is known as " secentismo" from its prevalence in the seventeenth century (Matteo Pellegrini, Orsi and others) felt the viciousness of the literary production of their day; they were aware that decadence was due to the fact that literature was no longer the serious expression of a content; but they were embarrassed by the reasoning of the champions of bad taste, who were able to demon

1 Diálogo de las lenguas (ed. Mayans y Siscar, Origines de la lengua española, Madrid, 1873), pp. 115, 119.

2 Essais, i. ch. 52 (ed. Garnier, i. 285); cf. ibid. chs. 10, 25, 39; ii. ch. 10.