Version classiqueVersion mobile
OpenEdition Books

Greek and Latin. Expressions of Meaning

 | 
Andreas T. Zanker

Chapter 8: The Classical Metaphors for Metaphor

Texte intégral

sed ea transferri oportet quae aut clariorem faciunt rem... aut quo significetur magis res tota siue facti alicuius siue consilii…

“But such metaphors should be used (a) as will either make the subject at hand more vivid... or (b) in order that the entire subject, whether it involves some deed or plan, should be better described” (Cicero, De Oratore 3.157).

  • 1 Besides Demandt (cited below), Derrida (1974), Ricoeur (1977/2006), 18-19, Kennedy (1991), 222 n. (...)

1We have seen how much of the language used by Greek critics to describe style came about by metaphor; Aristotle (Rhetoric 1405b34-1406b19), for example, notes that the overuse of compounds, outlandish words, epithets, and metaphors creates “frigidity” of style (“τὸ ψυχρόν”) and detracts from the “clarity” of writing (“τὸ σαφές”): these rhetorical terms were transferred from other domains. It is, however, possible to push things further: since Aristotle, it turns out, the western vocabulary for metaphor has itself been largely metaphorical. This is not anew observation.1 In the introduction to his book on metaphors for history, Alexander Demandt notes this feature of the trope with reference to Aristotle’s Poetics and Rhetoric:

1. “Neben dieser begrifflichen Klassifizierung von Metaphern beschreibt Aristoteles das Verhältnis der beiden Verwendungsweisen seinerseits in bildlicher Rede. Er gebraucht Metaphern für Metaphern”.

  • 2 Demandt (1978), 6.

“Besides this terminological classification of metaphors, Aristotle himself describes the relationship of the two modes of usage [i.e. metaphorical and non-metaphorical] in metaphorical language. He uses metaphors for metaphors”.2

2In the pages that follow, Demandt investigates a number of ancient meta-metaphors that continue to live on in the ways that we discuss lexical transferences today. One of these, attested since Aristotle’s Poetics and Rhetoric, arose from language for decoration. Metaphors were used, for example, “ad inlustrandam atque exornandam orationem” (“for the sake of illuminating and decorating speech” Cicero, De Oratore 3.152). Both Aristotle and Cicero describe metaphor in terms of clothing; according to the character of Crassus in Cicero’s De Oratore, just as clothing was first invented for the purpose of repelling cold but later came to be used for aesthetic reasons, so too metaphor, which was initially employed in order to create names for things, came to be used for the purpose of adornment (De Oratore 3.155; cf. Aristotle, Rhetoric 1405a13-14). A second group of metaphors for metaphor consisted of expressions such as “ὄνοµα οἰκεῖον” (“proper term”) and “ὄνοµα ἀλλότριον” (“borrowed/foreign term”) in Greek, “uerbum proprium” (“proper term”) and “uerbum alienum” (“borrowed/foreign term”) in Latin. The nouns associated with this last group of metaphors – for example, the term “proprietas” (“proper meaning”) – contributed to the ancient vocabulary of meaning.

3In what follows, I would like (I) to discuss the spatial metaphor implicit both in the term “metaphor” and in the ancient descriptions of the trope, before (II) considering the words used by the ancient philosophers and rhetoricians to describe metaphorical and non-metaphorical usages of language. One of my aims will be to demonstrate that the ancient metaphors for metaphor exemplified three functions of the trope mentioned by the ancient theorists: (a) its ability to provide a name for a hitherto unnamed thing, (b) the way in which it made vivid otherwise abstract concepts, and (c) its role in adorning prose and poetry. I shall then (III) move on to pose the question as to whether certain ancient theorists (Aristotle, Cicero, and Horace) reveal a consciousness of the metaphorical nature of their descriptions of metaphor, before (IV) briefly considering some of the metaphors employed for metaphor in the twentieth (and twenty-first) century, when a growing mistrust of the old terminology prompted the creation of new terms and modes of presentation.

  • 3 For a general history of metaphor, see the survey of Eggs in the HWRh s.v. “Metapher”. For the cla (...)
  • 4 Derrida (1974); Lloyd (1987), 172-214; Lloyd (1990), 14-38; Lloyd (1996), 205-222. For a good summ (...)
  • 5 My general account is influenced by views of language associated in particular with the work of Ge (...)

4The bibliography on metaphor is vast,3 yet the recursive aspect of the vocabulary – the fact that the word “metaphor” is itself a member of the set of words that it names – has only drawn sustained attention relatively recently. Perhaps the most important forerunners for my account are Derrida’s White Mythology, which sets the problem of metaphor at the center of philosophical inquiry, and G. E. R. Lloyd’s studies of metaphor in Aristotle and elsewhere.4 The aim of this chapter is not, however, to recapitulate their findings but to map the role of metaphor in the creation of the classical vocabulary for metaphor more broadly. The subject is of key importance to the book as a whole, given my emphasis on the role of metaphor in the creation of the classical vocabulary of meaning (Chapters 4and 5). I have argued that much of this vocabulary arose by metaphorical transference; to establish that the vocabulary of metaphor is itself the result of the process that it names would be to offer strong corroboration of metaphor’s power and ubiquity. In addition, the way in which the abstract concept of metaphor is described by terms that draw upon our embodied experience of the world further reinforces the directionality of the transference (“I-meaning” to “S-meaning”) that forms the backbone of my argument throughout the book.5

I. The Spatial Metaphor of “µεταφορά” and “translatio

  • 6 On Plato’s metaphors, see Pender (2000); she argues that the term “εἰκών” (“image”) may have at ti (...)
  • 7 See Stanford (1936), 3. Isocrates employs the term “µεταφορά” in anon-technical sense; poets, as o (...)
  • 8 On the “µετά” prefix as denoting “change”, see below.
  • 9 See Stanford (1936), 6. For metaphor as a process, see excerpt 2. For the use of “metaphor” to des (...)

5While Plato made extensive use of metaphor in his writing, he never employed the term “µεταφορά” itself.6 The word first appears in the fourth century BC in the Euagoras of Isocrates (9);7 it literally means “the carrying across” of a word from one thing to another,8 and could refer both to the process of transference (the phenomenon of metaphor in general) and to the transferred expression itself (a metaphorical word or phrase).9 The verb “µεταφέρειν” (“to carry across”) came to correspond with this usage of “µεταφορά”. The definition of metaphor in Aristotle’s Poetics has been foundational for all subsequent discussions:

2. µεταφορὰ δέ ἐστιν ὀνόµατος ἀλλοτρίου ἐπιφορὰ ἢ ἀπὸ τοῦ γένους ἐπὶ εἶδος ἢ ἀπὸ τοῦ εἴδους ἐπὶ τὸ γένος ἢ ἀπὸ τοῦ εἴδους ἐπὶ εἶδος ἢ κατὰ τὸ ἀνάλογον.

Metaphor is the application of a word that belongs to another thing; it occurs either from genus to species, from species to genus, from species to species, or by analogy” (Aristotle, Poetics 1457b6-9).

  • 10 See Kirby (1997); Crittenden (2003). The term “ἐπιφορά” already appears in the sense of “the appli (...)
  • 11 For Aristotle, this was a particular feature of metaphors involving ἐνέργεια (“activity”, “energy” (...)

6Here,“µεταφορά” (“carrying across”) is defined in terms of a similar word, “ἐπιφορά” (“setting next to”), which is perhaps best translated as “application”;10 in each case, a word for carrying (“φέρειν”) serves as a basis. The rather abstract notion involved in transferring words is thus described in terms of a concrete, dynamic action that we perform every day – that of carrying something from one place to another. The terminology preserves one of the functions of metaphor, or at least of a certain kind of metaphor, noted by the ancient commentators – that of putting something “πρὸ ὀµµάτων” or “ante oculos” (“before the eyes” Aristotle, Rhetoric 1411a26, Rhetorica ad Herennium 4.45).11

7It was not, of course, simply a matter of nouns and verbs. The metaphor of “carrying across” was reflected in the prepositions, adjectives, and adverbs employed by Aristotle, a scan be seen in the use of the preposition “ἀπό” in the following example:

3. δεῖ δὲ µεταφέρειν, καθάπερ εἴρηται πρότερον, ἀπὸ οἰκείων καὶ µὴ φανερῶν...

“It is necessary to draw a metaphor, just as we said previously, from things that are related to the object, but not too obvious” (Aristotle, Rhetoric 1412a10-11).

  • 12 Cf. “λέγω δὲ ἀπὸ γένους µὲν ἐπὶ εἶδος οἷον...” (“I suggest as an example of [a transference] from (...)
  • 13 Cf. “οὐ πόρρωθεν δεῖ ἀλλ᾽ ἐκ τῶν συγγενῶν καὶ τῶν ὁµοειδῶν µεταφέρειν” (“it is necessary to transf (...)

8In using metaphor, one carries a word “from” (“ἀπό”) one object “to” (“ἐπί”) a different one.12 The vividness of the above passage (3) is increased by the adjectival “µὴ φανερῶν” – in selecting a word to transfer, Aristotle suggests that it is best to go for the ones that are concealed (the metaphor is visual but implies spatiality) and are therefore likely to be novel. Indeed, Aristotle elsewhere claims that a skill with metaphor lies in “τὸ τὸ ὅµοιον θεωρεῖν” (“seeing similarity [between things]” Poetics 1459a8), where once again the stress is laid on the author’s ability to “visually observe” the various options. The spatial metaphor is also reflected in the adverbs used; Aristotle suggests that metaphors should not be carried across “πόρρωθεν” (“from far away”).13 The overall picture is ornate yet highly visualizable: authors have a range of words and things arrayed in front of them and should select which word to use for a thing based on their ability to spot similarities; they should aim to transfer words that are difficult to see, yet refrain from transferring those that are too far-fetched.

  • 14 OLD s.v. “translatio” 4: “transferred or figurative use (of a word)”; “translaticius” 1: “(of words (...)

9We see similar phenomena in Roman discussion of metaphor, where the Latin translation for “µεταφορά”, “translatio” (or “tralatio”), conserves the metaphor of “carrying across” found in the Greek; it is already found in the early first century BC in the Rhetorica ad Herennium (4.45).14 Extensions of the metaphor can be noted throughout the Roman descriptions of the trope. In the following excerpt from Cicero’s De Oratore, for example, Crassus describes metaphor in metaphorical terms:

4. uerbis... quae transferuntur et quasi alieno in loco conlocantur...

“Words... which are transferred and are, so to speak, situated in a foreign location” (Cicero, De Oratore 3.149).

10We will return to the term “quasi” in section III, but the main thing to emphasize at this point is the metaphor reflected in Cicero’s choice of verbs; transferred words are “set” (“conlocantur”) in a “foreign location” after they have been “carried over” (“transferuntur”) from their natural one. Elsewhere in Cicero’s rhetorical works one can find further extensions of the metaphor of moving words from one position to another, for instance the idea of “taking” a word out of its natural context in order to fill a gap in the vocabulary:

5. ergo hae translationes quasi mutuationes sunt, cum quod non habeas aliunde sumas.

“Consequently these metaphors, in which you take from elsewhere that which you do not have, are a kind of borrowing” (Cicero, De Oratore 3.156).

  • 15 This possibly engages with the old metaphor that equates words and coins; as noted, the word “uale (...)
  • 16 OLD s.v. “mutuor” b.

11When one takes something from somewhere else, it is “as if” (“quasi”) one were borrowing it. Here, there is a further refinement of the spatial metaphor – that of “loaning” (“mutuationes”).15 This extension was supported by the adjectives “proprius” (“one’s own”, “belonging to oneself”) and “alienus” (“foreign”, “belonging to another”) as used of supposedly proper and improper terms. The verb “mutuari” is used by Cicero in a similar manner in the following excerpt,16 where the use of the preposition “ab” is reminiscent of that found in the Aristotelian passage cited above (“ἀπό”):

6. ...necessitas cogat aut nouum facere uerbum aut asimili mutuari.

  • 17 Cf. Festus 136.23-25 Lindsay: “‘metaphoram’ quam Graeci uocant, nos ‘tralationem’, id est domo mut (...)

“...necessity forces us either to invent a new term or to borrow it from something similar” (Cicero, Orator 211).17

12To sum up, the very operation described in the two previous excerpts – transferring a word in order to create vocabulary for an object that lacks a name – requires the use of older vocabulary in new senses (“translatio”, “mutuatio”). In addition, the spatial metaphor is reinforced by means of adverbs and prepositions (“aliunde”, “ab”).

13Besides providing names for hitherto unnamed things, the application of a metaphor could also give pleasure – and here the spatial metaphor is similarly in evidence:

7. id accidere credo uel quod ingeni specimen est quoddam transilire ante pedes posita et alia longe repetita sumere; uel quod isqui audit alio ducitur cogitatione neque tamen aberrat, quae maxima est delectatio...

“I suppose that this [i.e. pleasure] occurs either because it is a mark of cleverness to jump over things that are set in front of our feet and to take things that are sought from afar, or because the listener is led somewhere else in his thought and yet does not go astray, which is a very great pleasure...” (Cicero, De Oratore 3.160).

  • 18 See Mankin (2011), 247.

14In this last excerpt a great deal of the description is couched in terms of physical position and movement. The scenario in Cicero is similar to what we saw in Aristotle: at the point of choosing a word, the author or speaker is conceived of as situated in afield populated by words and things. He should “jump over” the words in plain view and take ones that are less close-to-hand (the adverb “longe” plays the same role as “πόρρωθεν” in Aristotle). Beyond the metaphor of selecting words arrayed before one, we might note the spatial language in the description of the effect on the listener: he is “taken somewhere else” (“alio ducitur”). In the Orator (134), Cicero describes this “transporting” effect of metaphors on the mind of the receiver as in itself pleasurable.18

  • 19 See Quintilian, Institutio Oratoria 8.6.11: “praecipueque ex his oritur mira sublimitas, quae auda (...)
  • 20 Cited byInnes (2003), 7.

15At times, authors such as Cicero could further increase the vividness of their descriptions of metaphor by turning a metaphor into an active agent – personification, according to the term’s modern meaning. Aristotle had held this type of transfer (from animate to inanimate) in high esteem, as it was one way of endowing a depiction with “activity” (“ἐνέργεια”), and Quintilian considered the provision of agency to inanimate objects particularly vivid and powerful in descriptions.19 Cicero, for instance, could note that it was a feature of successful metaphorical transferences “...ut ea non inruisse in alienum locum, sed immigrasse in suum diceres” (“...that you would say they had not rushed into a foreign place but had migrated into their own” Cicero, Brutus 274),20 where words are described as agents moving through space. The following example comes from the orator’s letters, where the word “faithfully” is described as changing residence frequently:

8. unde in istum locum “fideliter” uenit? cui uerbo domicilium est proprium inofficio, migrationes in alienum multae.

  • 21 Cf. “etenim uerecunda debet esse translatio, ut deducta esse in alienum locum, non inrupisse, atqu (...)

From where does this ‘faithfully’ come into that place? The proper home for this word is in performing a duty, but it makes many incursions into a foreign setting” (Cicero, Ad Familiares 16.17.1).21

16The term is no longer “transferred” by an author but literally travels, seemingly of its own accord, from its “home” to a foreign one. Where as such terms had been set (“conlocantur”) in a foreign location in excerpt 4, here the metaphor itself is given agency.

  • 22 See Seneca, Epistulae 59.6; 114.10. On metaphor in Seneca, see Armisen-Marchetti (1989), (1991), ( (...)

17Cicero was not alone in this. Seneca could characterize metaphors by using words such as “audax” (“bold”) or “temerarius” (“rash”), where the daring of the individual employing the metaphor becomes attached to the metaphor itself.22 The personification of metaphors can be clearly seen in the following passage from Quintilian:

9. metaphora enim aut uacantem locum occupare debet aut, si in alienum uenit, plus ualere eo, quod expellit.

“For a metaphor ought to occupy an empty place, or, if it enters a foreign one, to be more impressive than that which it pushes out” (Quintilian, Institutio Oratoria 8.6.18).

  • 23 Cf. Seneca, Epistulae 58.1: “mille res inciderunt, cum forte de Platone loqueremur, quae nomina de (...)

18Words are described as material objects that can interact with one another, seemingly with intent and agency. A metaphorical item of vocabulary invades (“occupare”, “uenit”) foreign space, and expels the term to which it proves superior (“ualere”, “expellit”). The “locus” (“place”) that it occupies is figured as the spatial equivalent of the “res” (“thing”, “subject”) with which words correspond. At other times, the personification could be reversed: the things that required words (but lacked them) could be described as “desiring” them.23

  • 24 Cf. “τοῖς δ᾽ ὄνοµα ἄνθρωποι κατέθεντ᾽ ἐπίσηµον ἑκάστῳ” (“and for these things, human beings have a (...)
  • 25 Seneca also mentions the role of metaphor in setting something before the eyes; for similar statem (...)

19The language for metaphor interacted with more general ways of talking about the relationship between words and things. In both Greek and Latin descriptions of language words could be conceived of as positioned on top of the things that they described. For example, we read that metaphor consists of transferring “τὰ ἐπ᾽ ἄλλων κείµενα ὀνόµατα” (“words lying on top of [i.e. assigned to] other things” Dexippus, On Categories 6.13 Busse). This conceptualization is old and attested in Presocratic discussion of language.24 The idea that the name went “on top of” objects is also apparent in Latin: in the following excerpt, for example, Seneca is describing the attempt to translate the Greek word “σοφίσµατα” into Latin:25

10. quid uocentur Latine sophismata, quaesisti ame. multi temptauerunt illis nomen inponere, nullum haesit... aptissimum tamen uidetur mihi, quo Cicero usus est: “cauillationes” uocat.

  • 26 Cf. “...quibus etiam uerba parienda sunt inponendaque noua rebus nouis nomina” (“...by whom words (...)

“You have asked me what sophismata are called in Latin. Many have attempted to set a name on them, but none has stuck... The most fitting one, however, seems to be the one Cicero used: he called them ‘cauillationes’” (Seneca, Epistulae 111.1).26

  • 27 Compare Hermogenes’ usage of the verb “τιθέναι” (“to set”) for the “placement” of a word: “τροπὴ δ (...)
  • 28 Compare the vocabulary of the following passage: “δεῖ δὲ καὶ τὰ ἐπίθετα καὶ τὰς µεταφορὰς ἁρµοττού (...)

20Here, the name is literally “placed on” (“inponere”) a thing that lacks a Latin name;27 that previous attempts have not “stuck” (“haesit”) once again supplies a vivid mode of description – to put it somewhat comically, it appears that they have fallen off. Seneca’s application of “aptissimus” (“most fitting”) incidentally makes use of the same metaphor –“cauillatio” (“raillery”) “fits” best the thing it describes.28 This way of conceiving of the way in which names are placed “on top of” things is reflected in excerpts such as the following – the first (11) stemming from Epicurus’ Letter to Herodotus, preserved in Diogenes Laertius:

11. τὰ ὑποτεταγµένα τοῖς φθόγγοις... δεῖ εἰληφέναι...

“It is necessary to grasp the things [i.e. meanings] set under words” (Diogenes Laertius, Lives of the Philosophers 10.37).

12. ...non intellegere interdum, quid sonet haec uox uoluptatis, id est quae res huic uoci subiciatur.

“...and yet he [i.e. Epicurus] sometimes does not understand what the word ‘pleasure’ means, that is to say the thing that is placed under the term” (Cicero, De Finibus 2.6).

21These excerpts describe the things “under” the words rather than the words set “on top” of things (note too the metaphor of grasping = understanding in “εἰληφέναι”), but the relationship remains the same: names serve as the label by which things are identified, and are therefore figuratively conceived of as lying on top of them. Once again, the ideas of vision, space, and manipulation are intertwined in order to present a vivid picture of an abstract relationship.

22The spatial metaphor was also productive in the creation of vocabulary for other grammatical concepts, and it may be useful to briefly mention three examples. (a) Besides metaphor, “circumitio” (“circumlocution”, literally “a going around”) served as one way of getting around the problem of things that lacked a name; in the following passage, the verb (“circumire”) is used:

13. res plurimae carent appellationibus, uteas necesse sit transferre aut circumire.

“Many things lack names, with the result that it is necessary either to transfer [names] or to express [the thing] by circumlocution” (Quintilian, Institutio Oratoria 12.10.34).

  • 29 LSJ s.v. “σύνθετος” 2: “compound”; cf. “πρῶτον µὲν οὖν τρόποι ὀνοµάτων εἰσὶ τρεῖς, ἁπλοῦς ἢ σύνθετο (...)

23One can thus “go around” (“circumire”) something for which there is no name rather than “carry a name across” (“transferre”); once again, a concrete action (walking around an object) yields a term for an abstract rhetorical concept. (b) One might also offer as examples certain ancient terms for the parts of speech, such as the noun “τὸ σύνθετον” (“compound word”): with compound words, two elements are “placed together” in order to form a new whole.29 (c) To give a final example, terms such as “ὁρίζειν”, “ἀφορίζειν”, “finire”, and “definire”, which meant “to bound” or “enclose”, served as verbs for “defining” a word; in doing so, one marked off the word’s “domain”. Metalinguistic vocabulary of this kind – language for talking about language – is not something that appeared out of nowhere: it depended on a series of transferences from a physical domain to an abstract one, largely structured by a metaphor in which the relationship between words and things was figured in terms of space and movement.

  • 30 See Lausberg (2008), §562; cf. Stanford (1936), 37-39; Barwick (1957), 90.
  • 31 Cf. “τὸ γὰρ ‘µυρίον’ ‘πολύ’ ἐστιν, ᾧ νῦν ἀντὶ τοῦ ‘πολλοῦ’ κέχρηται” (“and the poet has used [the (...)
  • 32 Cicero could simply employ the term “translatio” (“metaphor”) and its relatives for both operation (...)

24Another word that could perform some of the same functions as “µεταφορά” was “κατάχηρσις”; this originally meant “misuse”, but came to refer to a necessary metaphor (or metonymy) that provided a name for a hitherto unnamed thing; it was translated into Latin by the word “abusio”.30 The associated verbs were “καταχρῆσθαι” (“misuse”, “to abuse”) and “abuti”, which were employed in the sense of “to misapply”, “to use in the wrong sense”, “use catachrestically”. The word appears to be contrasted with the unmarked “χρῆσθαι” (“to use”); Aristotle, for example, could describe metaphor in terms of the poet’s simply “using” (“κέχρηται”) one term “instead of” (“ἀντί”) another.31 The author of the Rhetorica ad Herennium (4.45) describes catachresis as a form of transference that consists in exchanging a proper word for a thing with one that is similar but improper. While the author of the Rhetorica took it as a natural enough feature of language, other writers were more positive in their descriptions, emphasizing its productivity in the creation of new vocabulary. The Alexandrian grammarian Tryphon (On Tropes, 192.21 Spengel) offered examples such as “χεῖλος κεραµίου” (“lip of ajar”), whereby a word properly employed of human beings comes to be used of inanimate objects, while Quintilian (Institutio Oratoria 8.2.4-5) referred to the catachrestic use of “lapidare” (literally, “to stone”) in order to provide a name for the throwing of clods and potsherds. According to a number of authors, metaphor and catachresis were to be distinguished: a metaphor consisted of the transference of an exotic name to an object that already possessed a name, whereas “κατάχηρσις”/“abusio” provided names for things that were as yet unnamed.32 The classic statement is in Quintilian:

14. eo magis necessaria κατάχρησις, quam recte dicimus “abusionem”, quae non habentibus nomen suum accommodat quod in proximo est...; discernendumque est ab hoc totum translationis istud genus, quod abusio est ubi nomen defuit, translatio ubi aliud fuit.

“Catachresis, then, which we with due right call by the name ‘abusio’, and which provides a nearby name for something that lacks its own, is all the more necessary...; this entire category [i.e. catachresis] is to be separated from metaphor, since catachresis is when a name [for the thing that receives the name] is missing, metaphor is when it is present” (Quintilian, Institutio Oratoria 8.6.34).

25The most important thing for us to note is that the word “κατάχηρσις” in its rhetorical sense was itself a catachresis – it came to mean “necessary metaphor” via a transference from an earlier sense.

  • 33 Bettini (2012), 35. The other Greek terms for “to translate” were “ἑρµηνεύειν” and “µεθερµηνεύειν” (...)

26Since we have broached the topic of translation, it is perhaps worthwhile mentioning that the terms “µταφέρειν” and “transferre” had more than one meaning when used with reference to language. Maurizio Bettini points out that these terms could be used not only in the sense of (a) “to transfer metaphorically”, but also in that of (b) “to translate”. Here are instances of the latter usage:33

15. Σόλων,... διαπυνθανόµενος τὴν τῶν ὀνοµάτων δύναµιν, ηὗρεν τούς τε Αἰγυπτίους τοὺς πρώτους ἐκείνους αὐτὰ γραψαµένους εἰς τὴν αὑτῶν φωνὴν µετενηνοχότας...

“Solon,... having investigated the meaning of the names, found out that the Egyptians who had first written them down had translated them into their own language” (Plato, Critias 113a).

16. istum ego locum totidem uerbis a Dicaearcho transtuli.

“I have translated this passage verbatim from Dicaearchus” (Cicero, Ad Atticum 6.2.3).

  • 34 The use of “µεταφέρειν” in the sense of “to change” was an old one; see, for example, Sophocles, P (...)
  • 35 The relevant Romance verbs stem not from “transferre” but from “traducere” (“to lead across”). For (...)

27As can be seen from the prior excerpt (15), the usage of “µεταφέρειν” in the sense of “to translate” is in fact attested quite early. Despite their being homonyms, the difference between the two operations (transferring and translating) is clear. With metaphorical transference, the form of the word in question remains static (although its meaning changes), whereas in the case of translation the forms of the words are “changed” or “transformed”; it is their meaning that is preserved.34 The two usages of “µεταφέρειν” thus depend on different ideas and metaphors. The Latin application of “translatio” in the sense of “translation” is once again simply a translation of the Greek equivalent, and through this the ambiguity of the Greek expression is preserved. The term “translatio” has in turn been calqued in German (“Übersetzung”) and borrowed in English (“translation”).35

ΙΙ. Proper and Improper Meanings of Words

  • 36 On this, see Derrida (1974), 48-49; Lloyd (1996), 205-222.
  • 37 See Barwick (1957). On Aristotle’s role in creating the categories “natural” and “deviant”, see Ll (...)
  • 38 Lausberg (2008), §533. Compare also the terms “ἰδιωτικός” (“non-specialized”) and “communis” (“sha (...)

28Metaphor, for ancient authors, consisted in bringing a word whose standard meaning lay elsewhere across in order to describe a different object. This could occur either in order to give a name to a thing that lacked a term (a process often termed “κατάχρησις”/“abusio”), to set things “before the eyes” of the reader/listener, or in order to provide a certain luster to one’s style. Given the nature of our larger project, however, it will be useful to investigate the metaphors for metaphorical and non-metaphorical vocabulary as well, such as those found in the phrases “proper term” and “natural meaning”.36 According to the latter, words are attached to things by nature, and one implication of this is that transferred words (that is, borrowings or metaphors) are often conceived of as unnatural; as will be seen later, this has been of key importance in modern debates concerning metaphor, but the vocabulary had its roots in the ancient world.37 In Greek and Latin there were several metaphorical terms derived from spatial and social domains that could be applied in order to differentiate between “standard” and “transferred” words: “κύριος”, “οἰκεῖος”, “proprius”, and “natiuus” on the one hand, “ἄκυρος”, “ἀλλότριος”, “ξενικός”, “improprius”, “alienus”, and “tralaticius” on the other.38 Two functions of metaphor are once again clearly exemplified in these metaphorical usages – (a) the creation of words for new things and (b) the rendering of an abstract operation in terms that make it vivid and set it “before the eyes”.

  • 39 LSJ s.v. “ἔτυµος” II (“τὸ ἔτυµον”): “the true sense of a word according to its origin, its etymolog (...)

29As a first example, we might turn to the substantive “τὸ ἔτυµον” (“the true [meaning]”), a technical term derived from an adjective meaning “true” since Homer (in the Iliad and Odyssey it is a doublet for “ἐτεός”). According to ancient authors, the “ἔτυµον” of a word consisted in its actual, non-transferred signification.39 Here is an example from Athenaeus:

17. καλοῦσι δὲ καὶ τὰς µισθαρνούσας ἑταίρας καὶ τὸ ἐπὶ συνουσίαις µισθαρνεῖν ἑταιρεῖν, οὐκ ἔτι πρὸς τὸ ἔτυµον ἀναφέροντες, ἀλλὰ πρὸς τὸ εὐσχηµονέστερον...

“And they call women who work for pay ‘hetairai’, and earning money for sex ‘hetairein’, not appealing to the true sense [of the word] but rather to its euphemistic one” (Athenaeus, Scholars at Dinner 571d).

  • 40 That is, Isis and Osiris were named “ἀπό τινος ἐτύµου τεθείσης ἑκατέρας τῆς προσηγορίας ταύτης” (“ (...)

30According to Athenaeus, the true sense of the word “ἑταιρεῖν”, “to keep company with”, is at variance with the euphemistic usage that Athenaeus records, where the verb means “to earn money for sex”. On other occasions, the “ἔτυµον” could reflect a fundamental tie between the word and the object that it described; in his account of the Egyptian gods Isis and Osiris, for example, Diodorus Siculus remarks that these names stand in a proper relationship to the deities to which they are attached: “Osiris” means “many-eyed”, which is a fitting description of a sun god (Library 1.11.1-2).40 Another term for this was “ἡ ἐτυµότης”, which Plutarch (Table Talk 638e) used in a similar way: the “ἐτυµότης” (“true meaning”) of the word “πάλη” (“wrestling”) could arguably be traced back to “παλύνειν” (“to sprinkle” with dust).

  • 41 Cf. DELL s.v. “κύριος”. Plato uses the term of the name giver himself at Cratylus 389d in the sens (...)
  • 42 Lausberg (2008), §1236-1237; Cope & Sandys (1970), 3.18. Aristotle, Poetics 1457b3-4, defines “κύρ (...)

31Besides “τὸ ἔτυµον”, one of the most important terms in describing the natural meanings of words was “κύριος”. One basic meaning of this adjective (as found in Pindar and Aeschylus) is “to have power over”.41 The substantive “ὁ κύριος” was used in classical Greek of the master of a house; when applied to things, the term meant something like “authoritative” or “principal”, “main”. In rhetoric and grammar, however, the term came to be applied to words themselves in the sense of “proper”, “standard”, or “established” (as opposed to loan-words, metaphors, or otherwise strange terms).42 In the following passage, Aristotle is describing the power of exotic language in epic:

18. καὶ ἐπὶ τῆς γλώττης δὲ καὶ ἐπὶ τῶν µεταφορῶν καὶ ἐπὶ τῶν ἄλλων ἰδεῶν µετατιθεὶς ἄν τις τὰκύρια ὀνόµατα κατίδοι ὅτι ἀληθῆ λέγοµεν.

“And in the case of loan-words, metaphors, and the other forms [of exotic language], you would see that we are speaking the truth [about the power of certain words in epic poetry] if you were to substitute the standard terms” (Aristotle, Poetics 1458b17-19).

  • 43 Concerning “ξενικός”, cf. “ξενικὸν δὲ λέγω γλῶτταν καὶ µεταφορὰν καὶ ἐπέκτασιν καὶ πᾶν τὸ παρὰ τὸ (...)
  • 44 LSJ s.v. “κυριότης”.
  • 45 LSJ s.v. “κυριολεξία’; “κυριολογία”.
  • 46 Cf. “κυριολογία µὲν οὖν ἐστιν ἡ διὰ τῆς πρώτης θέσεως τῶν ὀνοµάτων τὰ πράγµατα σηµαίνουσα...” (“th (...)

32Foreign elements such as metaphor elevate language. The adjective “κύριος” could be contrasted with “ἄκυρος” (“improper”), “ἀλλότριος” (“foreign”, “belonging to another”), or “ξενικός” (“foreign”, “strange”), as well as the terms “µεταφορά” and “µεταφορικός” themselves.43 Towards the end of antiquity, the noun “κυριότης”, which meant “dominion” in Koine Greek, begins to appear in philosophical and grammatical treatises in the sense of “the proper, legitimate use of a term”.44 Further related terms, such as “κυριολεξία” and “κυριολογία” were also employed with a similar meaning,45 the latter standing in contrast to the term “τρόπος” (“metaphor”) in the writing of Tryphon, whereas Apollonius Dyscolus could use it in the sense of the “proper meaning of a word” in his grammatical writing.46

33Besides the vocabulary related to “κύριος”, further terminology was used in Aristotle’s writing, for example where the philosopher argues that metaphor should be employed alongside proper terms in the writing of prose:

19. πάντες γὰρ µεταφοραῖς διαλέγονται καὶ τοῖς οἰκείοις καὶ τοῖς κυρίοις.

  • 47 Cf. “τὸ δὲ κύριον καὶ τὸ οἰκεῖον καὶ µεταφορὰ µόνα χρήσιµα πρὸς τὴν τῶν ψιλῶν λόγων λέξιν” (“stand (...)

“For everyone speaks using both metaphors and proper and standard words” (Aristotle, Rhetoric 1404b34-35).47

  • 48 LSJ s.v. “οἰκεῖος” IV.3: “a word in its proper, literal sense, opp. metaphor”; cf. Cope & Sandys (1 (...)
  • 49 LSJ s.v. “οἰκειότης” II: “of words and phrases, proper sense”.

34Here, the term “κύριος” is used in conjunction with “οἰκεῖος” (“of the same house”, “belonging to oneself”), a term that could likewise be used of the natural senses of words;48 the two adjectives were essentially synonymous in this particular usage, and the antonyms for “οἰκεῖος” were much the same as for “κύριος”. Just like “κυριότης”, the noun “οἰκειότης” could mean “the proper sense of a word”.49

  • 50 As Cope & Sandys (1970), 3.19 point out, Roman authors could translate both “κύρια” and “οἰκεῖα” a (...)
  • 51 OLD s.v. “proprius” 8. The OLD provides related usages: “quae sunt hominum propria... scribam” (“I (...)

35Latin words were once again found for the Greek ones; on the one hand, the functions performed by the words derived from “κύριος” and “οἰκεῖος” were assumed by the Latin adjective “proprius”, its noun “proprietas”, its adverb “proprie”, and its antonym “improprius”.50 The term “proprius”, literally “belonging to oneself”, “proper”, can be found used of words in Varro and Cicero:51

20. non proprio nomine... sed tralaticio, uta “pede” nostro “pes” lecti...

“Not with its own name... but with a transferred one, just as from our ‘foot’ we speak of the ‘foot’ of abed...” (Varro, De Lingua Latina 6.55).

  • 52 OLD s.v. “proprietas” 2. For the adverb “proprie”, compare the following: “tertium genus est eorum (...)
  • 53 OLD s.v. “proprietas” 1.

36In this excerpt a name is considered to belong to a thing; in other cases, however, a name that belongs to something else (“nomen alienum”) is “carried across” (“transfertur”) in order to describe an object. The noun “proprietas” was used by the post-Augustan authors of words in the specialized sense of “the proper or specific meaning”,52 although it too acquired this only as a transferred signification – an earlier sense was “the special quality or character of a thing”.53 The notion of ownership implicit in the word “proprius” supported further metaphors for the transference of terms such as that of “borrowing”: we should set the rhetorical usage of “proprius” in conjunction with the terms “mutuari” (“to borrow”) and “mutuatio” (“borrowing”) found in excerpts 5 and 6. In the following excerpt from Seneca, we see the verb “commodare” (“to borrow”) coupled with both “proprius” and “alienus”:

21. ingens copia est rerum sine nomine, quas non propriis adpellationibus notamus, sed alienis commodatisque: pedem et nostrum dicimus et lecti et ueli et carminis.

“There is a massive amount of things that have no name that we label with terms that do not belong to them, but are foreign/another’s property and borrowed: we use the word ‘foot’ of our own foot, but also use it of a couch or a sail or a poem” (Seneca, De Beneficiis 2.34.2).

  • 54 Cf. “nam si res suum nomen et uocabulum proprium non habet…” (“for if a thing does not have its ow (...)

37The word “pes” (“foot”), already encountered in an earlier chapter, is a classic example of a word for a human body part that has accrued different meanings on account of metaphorical transference. The notion of ownership is also evident in the phrasing of the relationship between words and things in terms of things having their “own proper” (“suum... proprium”) name.54

38Roman authors had recourse to other vocabulary in order to point out the perceived distinction between standard and metaphorical usage. Cicero, for example, used the term “natiuus” for the “natural” meanings of words, “repertus” (“invented”) for usages that were either transferred from these original meanings or coined afresh:

22. et simplicia uerba partim natiua sunt, partim reperta. natiua ea, quae significata sunt sensu; reperta, quae ex his facta sunt et nouata...

“Also, simple words are partly natural, partly invented: natural words are those whose nature is indicated by their meaning, while invented words are those that are created from natural words or otherwise are coined...” (Cicero, De Partitione Oratoria 16).

  • 55 For this narrative, see the following paragraph and Appendix III.
  • 56 Cf. Mankin (2011), 235. The same thing can be noted in Varro: there are three types of words – “un (...)

39Quintilian would later cite this passage in his discussion of tropes, adding that “natural” (“natiuus”) words were the ones that were first coined by primitive man (Institutio Oratoria 8.3.36).55 Elsewhere, Cicero describes propria uocabula (“proper terms”) as being “paene una nata cum rebus ipsis” (“almost born at the same time as the things [that they describe] themselves” Cicero, De Oratore 3.149), where the genetic metaphor is preserved.56

  • 57 See Appendix III. The main texts for this aspect of Stoic thought on this topic are Origen, Contra (...)
  • 58 There is an irony in Quintilian’s opening statement on the relationship between lucidity and propr (...)

40These last metaphors contrast a “natural” and a “transferred” usage of words: words have a “proper”, “native” meaning. The vocabulary therefore fits in well with naturalist theories of language such as that of the Stoics.57 This has been one of the aspects that modern critics have attacked in classical discussions of metaphor, in that the demarcation points between natural and transferred usages are often less certain than the vocabulary suggests, but an unease with the dichotomy can already be marked in ancient engagement with the terminology. Early on in the tradition there was a tension between the two categories, a number of prominent authors recognizing the fact that apparently standard, non-poetic language was itself imbued with metaphor. Aristotle himself, as we have seen, noted that we all use metaphor in speech, which would seem to suggest that he intuited the problematic nature of his own disjunction between standard and non-standard usage; Cicero (Orator 81) and Quintilian (Institutio Oratoria 8.6.4) also stated that everyone uses metaphor, including farmers, while Seneca (Epistulae 59.6) mentioned that metaphors were not the preserve of poets, but were in fact used by the ancient prose writers.58

  • 59 On Demetrius’ discussion of metaphor, see Schenkeveld (1964), 88-99.

41According to the rhetorician Demetrius, metaphorical terms could in fact be more “proper” than the proper terms themselves:59

23. ἔνια µέντοι σαφέστερον ἐν ταῖς µεταφοραῖς λέγεται καὶ κυριώτερον, ἤπερ ἐν αὐτοῖς τοῖς κυρίοις, ὡς τὸ “ἔφριξεν δὲ µάχη”.

  • 60 Cf. “etiam quae bene translata sunt, propria dici solent” (“those words too that are transferred w (...)

“Some things, however, are mentioned more clearly and properly in metaphor than they are in the proper terms themselves, for example in the case of the phrase ‘the battle shuddered’” (Demetrius, On Style 82).60

  • 61 Note that throughout this description, Demetrius makes συνήθεια itself an active agent (“µεταφέρου (...)
  • 62 Schenkeveld (1964), 97, argues that Demetrius’ point is that “the κύρια, the literal, usual, prope (...)

42“The battle shuddered” is an example of personification (transference from animate to inanimate); Demetrius continues by noting that the use of the proper term instead of “shuddered” would not be more conducive of clarity with regard to the phrase’s meaning. Later on (On Style 86-87), Demetrius maintains that standard usage (“συνήθεια”) in fact expresses nearly everything by metaphor, and does so in such a secure way (“ἀσφαλῶς”) that we barely notice that the metaphorical terms are transferences. He provides several examples of such “naturalized” metaphors, for example “clear voice” and “harsh man” (“λευκήν τε φωνήν… καὶ ὀξὺν ἄνθρωπον”).61 Naturalized metaphors such as these serve the same functions as the proper words themselves, “ὅµοια δοκεῖν τοῖς κυρίοις” (“they seem the same as the proper ones”), to the point that in some cases we no longer need the latter – for instance in the case of the phrase “ὁ τῆς ἀµπέλου ὀφθαλµός” (“the eye/bud of a vine” On Style 87), a metaphorical transference that has apparently completely ousted the original term.62 While the dichotomy introduced by terms such as “κύριος”, “οἰκεῖος”, and their correlates has been influential in subsequent discussion of metaphor, and was strongly criticized by twentieth-century critics such as I. A. Richards, these categories were in fact already being questioned in the Classical period.

III. Ancient Awareness of Metaphors for Metaphor

43In the preceding pages I have considered (I) the way in which the Greek and Latin terms for “metaphor” are themselves metaphorical, and (II) how much of the terminology for metaphorical and non-metaphorical language arose by means of the same trope. It is clear that these factors could lead to the description of metaphor in quite striking images that went well beyond the basic terminology. Two examples follow:

24. καὶ πᾶσα τέχνη φων[ὴ]ν οὐ δύναται προ [ίεσ-
θαι στερ[η]θεῖσα τῆς ἐκ τῶν µεταφορῶν εὐχρησ[τίας.

“And every art is unable to utter a sound if it is deprived of service of metaphors” (Philodemus, On Rhetoric 1.175 Sudhaus).

25. πάντων δὲ καὶ τῶν ἄλλων ἡ συνήθεια καὶ µάλιστα µεταφορῶν διδάσκαλος.

Usage is our teacher in all things, but especially when it comes to metaphor” (Demetrius, On Style 86).

44Philodemus (24), while describing the power of metaphor, has “every art” pronouncing a voice, while Demetrius (25) claims that “usage” is the “teacher” when it comes to metaphor. In Latin, ornate metaphors for metaphor can be found in Seneca and Quintilian. In excerpt 26, Seneca is arguing that metaphor is not limited to poetry – even the early Latin prose authors made use of it:

26. illi, qui simpliciter et demonstrandae rei causa eloquebantur, parabolis referti sunt, quas existimo necessarias, non ex eadem causa qua poetis, sed ut imbecillitatis nostrae adminicula sint, ut et dicentem et audientem in rem praesentem adducant.

  • 63 My thanks to John Schafer for pointing this excerpt out to me. On this passage, see Bartsch (2009) (...)

“They [the ancient prose writers], who spoke simply and for the purpose of stating their case, are crammed full of comparisons; I judge these to be necessary, not for the same reason as the poets but so that they might be props for our weakness and might bring the speaker and listener face to face with the thing under discussion” (Seneca, Epistulae 59.6).63

45Not only do we have clear mention here of the role of metaphor in making the subject matter more vivid – we also see authors (author for text) “stuffed” with comparisons, which are likened to sticks or props (originally a farming metaphor). We might compare these physical metaphors with the metaphor of light found at the beginning of Quintilian’s treatment of metaphor in the Institutio Oratoria:

27. ...tum ita iucunda atque nitida utinoratione quamlibet clara proprio tamen lumine eluceat.

“...and it [metaphor] is so pleasant and resplendent that, no matter how vivid the language in which it is embedded, it shines forth with its own light” (Quintilian, Institutio Oratoria 8.6.4).

46Quintilian’s language in this excerpt apparently enacts precisely what it describes: the metaphor jumps out from the rest of the sentence. It is of course impossible to describe metaphor without metaphors, but such examples should give us reason to pause and to consider whether a certain irony is being applied – that is, whether the metaphors employed in excerpts such as these were deliberate.

  • 64 See Lloyd (1996), 213: “Aristotle does not explicitly draw attention to the fact that his whole ac (...)
  • 65 Cf. Aristotle, Posterior Analytics 97b37-39: “εἰ δὲ µὴ διαλέγεσθαι δεῖ µεταφοραῖς, δῆλον ὅτι οὐδ᾽ (...)

47In what follows, I would like to consider in greater depth how three ancient literary thinkers (Aristotle, Cicero, and Horace) may reveal a consciousness of the metaphorical quality of the vocabulary for metaphor. G. E. R. Lloyd, for example, has put forward precisely this possibility in the case of Aristotle’s Poetics and Rhetoric.64 Aristotle, who attacked the use of metaphor in scientific writing on account of its tendency to vitiate syllogisms and generally mislead,65 nevertheless freely used metaphors in his exposition of metaphor in these two works, and the question remains as to why this was so. Lloyd broaches the possibility that Aristotle classed his rhetorical writing itself under the heading of rhetoric, a field in which the philosopher in fact praises the skilled use of metaphor. The metaphors for metaphor used within the Rhetoric and Poetics would serve, then, to demonstrate the use of metaphor as it is described within the works themselves. The Rhetoric teaches the art of persuasion; therefore, Lloyd continues,

  • 66 Lloyd (1996), 215.

28. “...if Aristotle’s own treatment exemplifies that, being itself rhetorical in the sense that it applies what it teaches about the means of persuasion on any subject-matter (in this case the topic of rhetoric itself), then there would be nothing at all surprising about the use of metaphora in the account of it”.66

  • 67 Lloyd (1996), 217: “The encouragement to see the work as self-referential is minimal, even though (...)
  • 68 Lloyd (1996), 219-222. The relevant excerpt from Dexippus can be found at the beginning of Chapter (...)

48Despite the argument’s ingenuity, Lloyd remains skeptical, and he ultimately finds it too much of a stretch to see the Rhetoric as instantiating Aristotle’s own views about rhetoric.67 The Rhetoric does not, for example, fit easily into any of the three major categories of rhetoric according to Aristotle – deliberative, forensic, or epideictic. Furthermore, it must be conceded that metaphor appears in Aristotle’s works beyond the Rhetoric, including the logical and metaphysical treatises; there is no need to see the Rhetoric as a special case. In sum, Lloyd argues that there is a split between theory and practice in Aristotle’s writing when it comes to metaphor – in theory, Aristotle advocates doing away with metaphor in philosophical writing, but his own practice diverges from this ideal. As Dexippus pointed out, metaphor is an important force in philosophical discussion: Aristotle can simply not do without it.68

49There is, however, a further reason to think that Aristotle encourages his reader to note a self-referentiality and awareness of the metaphorical nature of his language for metaphor. In both the Poetics (1459a6-7) and Rhetoric (1405a9-10) he makes the point that a skill with metaphor is something that cannot be gained from anyone else – in the phrasing of the Poetics, “οὔτε παρ᾽ ἄλλου ἔστι λαβεῖν” (“nor is it possible to receive this from another”). The fact that a skill for “µεταφορά” (or “τὸ µεταφορικόν”), whose very name contains the idea of “carrying across”, cannot be transmitted from one person to another might be considered ironic, particularly in works that are designed to impart skill in rhetoric and poetry. One can transfer words between things, but one cannot transfer ability with metaphor from one human being to another. That Aristotle repeats the idea in very similar phrasing in both texts may encourage us to interpret it as pointed – to see a play on the constituent elements of the word “metaphor” in Aristotle’s statement that a skill with the trope cannot be “transferred”.

50The writing of Cicero on the topic offers a further possibility; the metaphors that I have been considering are clearly on display in the following passage from the De Oratore (I have already discussed parts of this excerpt):

29. ergo utimur uerbis aut iis, quae propria sunt et certa quasi uocabula rerum paene una nata cum rebus ipsis, aut iis, quae transferuntur et quasi alieno in loco conlocantur, aut iis, quae nouamus et facimus ipsi.

“Thus, the words that we use are, so to speak, either the proper and definite designations of things and were almost born at the same time as their objects themselves; or we employ those that are transferred and so to speak situated in a foreign location; or those that we ourselves create and coin” (Cicero, De Oratore 3.149).

  • 69 Cf. OLD s. v. “quasi” 9: “(used to tone down or apologize for a word or phr. used in an unusual wa (...)
  • 70 Mankin (2011), 235, takes “quasi” here with the entire clause; the term “uocabula” hardly needs qu (...)

51Besides the other vocabulary that this excerpt contains, it is worth pointing out that the character of Crassus calls attention to his own figurative language by his use of the word “quasi” (“as if”, “so to speak”), a word frequently employed in the De Oratore in order to introduce a metaphor or other trope.69 The first group of words is “born at the same time as the objects themselves” only in a manner of speaking;70 Crassus points to the metaphorical status of “alieno in loco conlocantur” by means of the same word. To reinforce this usage, one might offer the following passage:

30. ...Phalereus Demetrius meo iudicio praestitit ceteris; cuius oratio cum sedate placideque liquitur tum illustrant eam quasi stellae quaedam tralata uerba atque mutata.

“Demetrius of Phalerum led them all, in my judgment. His oratory not only proceeds in a calm and peaceful flow, but is lit up by, so to speak, the stars of ‘transferred’ and ‘borrowed’ words” (Cicero, Orator 92).

52Once again, a metaphor for metaphor is introduced by “quasi”, which brings the transferred nature of the description to the fore: metaphors are stars that so to speak “make resplendent” (“illustrant”) the writing in which they are set. Finally, one might also note the use of “quasi” in excerpt 5, where Cicero describes metaphors “as if” they are borrowings.

53Cicero was certainly aware of the function of words such as “quasi” in introducing metaphors; in the following excerpt from De Oratore, he mentions the possibility of ameliorating the harshness of a metaphor by preceding it with a qualifying term or phrase:

31. atque etiam, si uereare ne paulo durior translatio esse uideatur, mollienda est praeposito saepe uerbo; ut si olim, M. Catone mortuo “pupillum” senatum quis relictum diceret, paulo durius; sin “ut ita dicam pupillum” aliquanto mitius.

“And if you should fear that a metaphor should appear rather harsh, it should be softened with a word set in front, as is frequently done; as it might seem rather harsh if, when Marcus Cato was dead, someone were to have said that the senate had been left ‘an orphan’, whereas it would have been a little milder if he had said ‘what I might call an orphan’” (Cicero, De Oratore 3.165).

54Cicero here uses the phrase “ut ita dicam”, which performs much the same function as “quasi”: terms such as these soften otherwise stark metaphors. The word “quasi” figures prominently in Cicero’s discussions of metaphor, and this may suggest that he was aware of the figurative nature of the language he was employing to describe it.

  • 71 On this absence, see Innes (2003).

55An awareness of the metaphorical origins of the vocabulary may also be suspected in the writing of Horace, an author who, despite his mastery of metaphor, strangely avoids the subject in his Ars Poetica.71 As has been noted, Roman authors from the first century BC onwards had at their disposal terms such as “proprius” and “natiuus” in referring to “proper” or “natural” meanings as opposed to metaphorical ones. Nevertheless, Horace introduces a further item in his Ars Poetica, translating “κύριος” by means of the participle “dominans”:

32. non ego inornata et dominantia nomina solum
uerbaque, Pisones, Satyrorum scriptor amabo...

“I shall not cling only to words that are unadorned and used in their dominant sense, o Pisos, if writing satyr plays” (Horace, Ars Poetica 234-235).

  • 72 Brink (1971), 285-286. Cf. Rudd (1989), 189: “By using dominantia instead of the usual propria, H. (...)
  • 73 For similar metaphors in Horace, replacing theory with practice, see Innes (2003), 21-27: “the the (...)

56C. O. Brink points out that this usage of “dominans” is practically unparalleled in classical literature and may well be Horace’s own innovation; he further suggests that the use of such a word is pointed in this context – Horace apparently exemplifies the key idea of the passage within its very language, using a word in a transferred sense in order to articulate the need for precisely such words in satyr drama.72 The adjective “dominans” is not, in fact, the proper term for what it describes in this context, but neatly comments on the attitude towards language that Horace advocates for satyr plays. It would have been easy for the poet to use the phrase “propria uerba” instead of “dominantia uerba”.73 In fact, Horace might be thought to improve on the terminology, or at least to bring it in line with the metaphor of the Greek “κύριος”. While synonymous with “proprium” in this excerpt, “dominans” is in fact a direct translation of the Greek word: both “dominans” and “κύριος” mean “having power over”. In introducing this new item of vocabulary, Horace may additionally be making a point about the deficiencies of the “proper” item of vocabulary.

57None of these examples quite represents a “smoking gun” for a consciousness of meta-metaphor: the usage of metaphors by an author known to criticize metaphor in scientific writing, such as Aristotle, may simply reveal a minor blind-spot in his reflections on language; how often do we ourselves actively consider the metaphors that we use while writing? Similarly, Cicero’s use of “quasi” in order to introduce his metaphors for metaphor may merely reflect his general modus operandi and not any particular awareness of this feature of his language for metaphor. And while Brink’s argument concerning “dominans” in Horace is neat, the Augustan poet’s usage may likewise be unintentional. Nevertheless, the possibility of an awareness of the transferred nature of the vocabulary of metaphor should not be dismissed out of hand, particularly in the case of these three authors.

IV. Modern Metaphors

  • 74 For an introduction to modern theories of metaphor, see Pender (2000), 3-27.

58What of the modern vocabulary for metaphor?74 The words for “metaphor” have been assimilated into the modern languages either by naturalization (for example, English “metaphor”, “transference”, and German “Metapher”) or calque (German “Übertragung”), as has the terminology for the “natural”, “dominant”, and “proper” usage of words (not to mention “foreign”, “farfetched”, or “borrowed” expressions). One might also note continuity when it comes to the use of prepositions in such descriptions (we still speak of transferring words “to” and “from” things). Nevertheless, new ways of conceptualizing metaphor have encouraged the application of new vocabulary. The introduction of the terms “tenor” and “vehicle” by I. A. Richards is a case in point. In The Philosophy of Rhetoric, Richards criticized what he saw as the confused and demoralized state of rhetoric in the modern age. In Richards’ view, part of this resulted from the belief that there were “proper” or “natural” meanings to words (an idea supported by the classical vocabulary):

  • 75 Richards (1936/2001), 6-7.

33. “A chief cause of misunderstanding, I shall argue later, is the Proper Meaning Superstition. That is, the common belief – encouraged officially by what lingers on in the school manuals as Rhetoric – that a word has a meaning of its own (ideally, only one) independent of and controlling its use and the purpose for which it should be uttered”.75

  • 76 For a lucid account, see Ricoeur (1977/2006), 89; cf. Silk (1974), 6-14.
  • 77 This is also the view of Stanford (1936), 8, who cites Augustine, De Dialectica 9.

59Richards pointed out that words change their meanings according to the sentences in which they are embedded – it is only when one looks to the context in which words are housed that their meanings become clear.76 He argued for a “context theorem of meaning” that dispenses with the distinction between “natural” and “foreign” usage: words do not have a “proper” meaning in and of themselves, but rely on their context for their determination.77

  • 78 Richards (1936/2001), 63: “Thought is metaphoric, and proceeds by comparison, and the metaphors of (...)
  • 79 Richards (1936/2001), 64.
  • 80 Richards (1936/2001), 64.

60In the fifth and sixth lectures of The Philosophy of Rhetoric, Richards turned to the issue of metaphor, which he saw as all-pervasive and, prefiguring the cognitive approach of the late twentieth century, as rooted in thought rather than in language alone.78 According to Richards, rhetoricians had hitherto lacked a terminology for “distinguishing from one another what Dr Johnson called the two ideas that any metaphor, at its simplest, gives us”.79 Richards dubbed these the “tenor” and “vehicle” – the first term referring to the “underlying idea or principal subject which the vehicle or figure means”, the second referring to the item that “conveys” the tenor.80 The word “metaphor” is used to refer to the combination of this double unit. The relationship can be depicted (via a visual metaphor) as follows:

61By means of this terminology, Richards could do away with clumsy oppositions such as “the original idea” and “the borrowed one”, “the principal subject” and “what it resembles”, or “the idea” and “its image”.

62It is unclear where the vocabulary of “tenor” and “vehicle” came from. Richards may be using the term “tenor” in the same sense as Berkeley in a passage that Richards quotes earlier in The Philosophy of Rhetoric. As Ricoeur observes:

  • 81 Ricoeur (1977/2006), 398 n. 27.

34. “The fundamental meaning of the term tenor is captured in the following passage by Berkeley, quoted by Richards (4-5): ‘I do... once for all desire whoever shall think it worth his while to understand... that he would not stick on this or that phrase, or manner of expression, but candidly collect my meaning from the whole sum and tenor of my discourse, and laying aside the words as much as possible, consider the base notions themselves...’”.81

  • 82 Cf. Ricoeur (1977/2006), 93. See, however, Silk (1974), 12.
  • 83 Silk (1974), 13, argues for the usefulness of the terms.

63In this case, Richards would be using “tenor” in the sense of “the general meaning”. Richards’ use of “vehicle” is even less clear: it may be that the vehicle is conceived of as “carrying” or “conveying” the tenor within it, but the terminology is at the very least, to use Ricoeur’s terminology, “esoteric”.82 The fact that Richards’ expressions “tenor” and “vehicle” do not make clear use of the spatial metaphor elsewhere employed in the tradition results in their being difficult to memorize and keep straight in one’s mind: the terminology lacks the virtue attributed to metaphor by the classical rhetoricians – that of putting the thing described “before the eyes”. Richards is of course attempting to get away from the old metaphors, but the terms that he adopts are undeniably less user friendly than those that they supplant.83

  • 84 “To speak of the ‘interaction’ of two thoughts ‘active together’ (or, again, of their ‘interillumi (...)
  • 85 “Let us try, for instance, to think of a metaphor as a filter”; Black (1954-1955), 286.
  • 86 Black (1954-1955), 288.

64Subsequent investigations of metaphor have returned to the use of overt metaphors, often self-consciously so; in a seminal article, the philosopher Max Black attempted to get away from the standard “substitution” theories of metaphor (supported by the classical terminology) by introducing his theory of “interaction”:84 the expression “man is a wolf”, for example, involves the term “wolf” as a “filter” that allows us to map certain aspects of wolves onto human beings.85 The metaphor emphasizes the wolf-like traits of men and suppresses the non-wolf-like ones: “in short, it organizes our view of man”.86 Besides the metaphor implied by the term “filter”, Black also introduced the terms “focus” and “frame”. In the sentence “the chairman ploughed through the discussion”, for example, Black comments:

  • 87 Black (1954-1955), 276.

35. “Let us call the word ‘ploughed’ the focus of the metaphor, and the remainder of the sentence in which that word occurs the frame.(Are we now using metaphors – and mixed ones at that? Does it matter?)”.87

65Black is aware that he is using metaphors to employ metaphors, but evidently finds the terminology a convenient means of describing what he wants to say. Once again, metaphor serves as an important tool in the discussion of metaphor.

  • 88 This terminology is used by Lakoff & Johnson (1980), for example.
  • 89 For example, Aristotle’s example of Gorgias’ cry “Fie Philomela!” when a nightingale excreted on h (...)
  • 90 Fauconnier & Turner (2003), 40-44.

66To close, one might compare the terms of choice for contemporary cognitive linguists – “source” and “target” – which, even if the metaphor is once again “mixed”, nevertheless preserve the sense of movement through space found within the classical tradition; the idea is also immanent in current terms such as “domain”, “mapping”, and so on.88 Recent discussions of metaphor in cognitive linguistics have also turned to visual representations of the phenomenon; these are, in effect, visual metaphors for metaphor. In the work of Fauconnier and Turner on “cognitive blending”, for example, which attempts to account for emergent features of a metaphor not found in either the source or target domains,89 we see representations of the process that visually depict input spaces, cross-space mapping, generic space, and the blend itself.90 In the visual representation of blending, the phenomenon is quite literally placed before the eyes (“πρὸ ὀµµάτων”, “ante oculos”). The spatial metaphor for metaphor, reflected in ancient descriptions of the trope as well as in the very word itself, is often portrayed by contemporary cognitive linguists in terms of space on the page.

V. Conclusion

67The classical vocabulary of metaphor is itself metaphorical; it thereby exemplifies certain functions of metaphor as described by the ancient rhetoricians (Aristotle, Demetrius, the author of the Rhetorica ad Herennium, Cicero, Quintilian, and so on). (a) Metaphorical (or catachrestic) transference provided names for a hitherto unnamed phenomenon (metaphor). Greek thinkers came to discuss metaphor at a relatively late date (even though metaphor had been a force in Greek poetry and prose from their very beginnings); in order to create terminology for the new idea under discussion, words were “brought across” from pre-existing lexical domains. (b) The sphere from which the new vocabulary of metaphor was transferred was largely spatial – “metaphor” was framed in terms of “carrying something across”, and subsidiary vocabulary was borrowed from the same domain. This had the result of making descriptions of metaphor vivid, setting the phenomenon “before the eyes” and thus easing comprehension. Finally, (c) ancient authors could use meta-metaphors in order to decorate their poetry and prose – another key function of the trope according to the rhetoricians. Cicero describes metaphor as if (“quasi”) it illuminates writing, while Quintilian invests his descriptions of metaphor with ἐνέργεια (“activity”) by turning the word “metaphora” itself into an active agent.

68There is no reason to see a substantial break between our own terminology for these processes and that used in antiquity – the modern vocabulary retains the metaphors that were first used in the ancient world, and in fact this shared link quite often blinds us to the metaphorical nature of what seem to us rather antiseptic and clinical terms. Indeed, these metaphors even retain a hold on the phrasal lexicons of those who neither know the classical languages nor have read the ancient discussions of metaphor in translation: when speakers of modern English talk about metaphorical transferences, the word “transference” still brings with it the association of “changing place”. It is in fact only by going back to the ancient authors that we can observe the genesis of the conceptual structures that continue to feature in modern discussions of thought and language. A sizable amount of the classical vocabulary of meaning evolved by the transference of verbs that described actions with which human beings are intimately acquainted (wanting, intending, and giving a sign); it turns out that the vocabulary of metaphor was also structured in terms of our lived experience of the world.

Notes

1 Besides Demandt (cited below), Derrida (1974), Ricoeur (1977/2006), 18-19, Kennedy (1991), 222 n. 25, Lloyd (1996), 211, Kirby (1997), Innes (2003), 7, and Crittenden (2003), to name just a few, have stressed the metaphor contained within the word “metaphor”.

2 Demandt (1978), 6.

3 For a general history of metaphor, see the survey of Eggs in the HWRh s.v. “Metapher”. For the classical material, see Lausberg (2008), § 558-564, and Novokhatko (2014), 415; the pioneering study on ancient metaphor is Stanford (1936). Key classical texts include Ps.-Aristotle, Rhetoric to Alexander 1434b34; Isocrates, Euagoras 9; Aristotle, Rhetoric 1404b32-1405b20, 1407a10-17, 1410b13, etc.; Poetics 1457b6-33; Demetrius, On Style 78-90; Cicero, De Oratore 3.149-169; Orator 81, 92-95, 202; Quintilian, Institutio Oratoria 8.6.4-18; Ps.-Longinus, On the Sublime 32; Hermogenes, On Invention 4.10; Isidore, Etymologiae 1.37.2-6; Bede, De Schematibus et Tropis 2. Cf. Calboli (1993), 391-393. For a more expansive selection (including fragments of Theophrastus, Philodemus, and the grammarians collected in Spengel) in a convenient format, see Guidorizzi & Beta (2000).

4 Derrida (1974); Lloyd (1987), 172-214; Lloyd (1990), 14-38; Lloyd (1996), 205-222. For a good summary of their arguments, see Crittenden (2003).

5 My general account is influenced by views of language associated in particular with the work of George Lakoff and his collaborators; see Lakoff & Johnson (1980); Lakoff (1987); Lakoff & Turner (1989); Ortony (1993); Lakoff & Johnson (1999). See Sweetser (1990), 18-19, on the standard directionality of catachrestic metaphor: “more abstract domains of meaning tend to derive their vocabulary from more concrete domains (rather than vice versa)”. Cf. Cicero, De Oratore 3.161: “illa uero oculorum multo acriora, quae paene ponunt in conspectu animi, quae cernere et uidere non possumus” (“the [metaphors] drawn from vision are much more vivid, since they almost place the things that we are unable to see ourselves within the sight of our mind”).

6 On Plato’s metaphors, see Pender (2000); she argues that the term “εἰκών” (“image”) may have at times functioned as a Platonic synonym for the later “µεταφορά”. This rhetorical term was, on the face of it, also a product of metaphor, as noted by McCall (1969), ix: “Almost all ancient technical terms were originally nontechnical in meaning... Thus, for example, before being used as a rhetorical term, “εἰκών” meant ‘statue’ or ‘portrait’ or ‘image’”.

7 See Stanford (1936), 3. Isocrates employs the term “µεταφορά” in anon-technical sense; poets, as opposed to prose authors, are able “...περὶ τούτων δηλῶσαι µὴ µόνον τοῖς τεταγµένοις ὀνόµασιν, ἀλλὰ τὰ µὲν ξένοις, τὰ δὲ καινοῖς, τὰ δὲ µεταφοραῖς, καὶ µηδὲν παραλιπεῖν, ἀλλὰ πᾶσι τοῖς εἴδεσι διαποικῖλαι τὴν ποίησιν” (“...to treat these subjects not only by means of conventional expressions, but in words now exotic, now newly coined, now in figures of speech, and to leave nothing out, but to use every kind that might decorate their poetry” Isocrates, Euagoras 9).

8 On the “µετά” prefix as denoting “change”, see below.

9 See Stanford (1936), 6. For metaphor as a process, see excerpt 2. For the use of “metaphor” to designate a word, compare: “καὶ ἐάν τε κοσµεῖν βούλῃ, ἀπὸ τῶν βελτίστων τῶν ἐν ταὐτῷ γένει φέρειν τὴν µεταφοράν...” (“and should you wish to ornament [your subject, it is necessary] to transfer a metaphor from the best species under the same genus” Aristotle, Rhetoric 1405a14-16).

10 See Kirby (1997); Crittenden (2003). The term “ἐπιφορά” already appears in the sense of “the application of names” in Plato, Cratylus 430d; Laws 944b: “ἡ τούτων τῶν ὀνοµάτων ἐπιφορά” (“the application of these names”).

11 For Aristotle, this was a particular feature of metaphors involving ἐνέργεια (“activity”, “energy”, “actuality”), such as personifications or descriptions of movement; see his definition of “πρὸ ὀµµάτων” at Rhetoric 1411b24-1412a9. For example, Aesion’s statement that the Athenians had drained their entire city into Sicily (a reference to the Athenian expedition) is described as a metaphor – one “before the eyes” (Aristotle, Rhetoric 1411a25-26).

12 Cf. “λέγω δὲ ἀπὸ γένους µὲν ἐπὶ εἶδος οἷον...” (“I suggest as an example of [a transference] from genus to species the following...” Aristotle, Poetics 1457b9-10). For similar use of prepositions, compare the definitions of Tryphon and Ps.-Plutarch: “μεταφορά ἐστι λέξις μεταφερομένη ἀπὸ τοῦ κυρίου ἐπὶ τὸ μὴ κύριον ἐμφάσεως ἢ ὁμοιώσεως ἕνεκα” (“metaphor is an expression that is transferred from its proper usage to a non-proper usage, either for the sake of emphasis or because of resemblance” Tryphon, On Tropes 191.23-192.1 Spengel); “μεταφορὰ δέ, ἥπερ ἐστὶν ἀπὸ τοῦ κυρίως δηλουμένου πράγματος ἐφ᾽ ἕτερον μετενηνεγμένη κατὰ τὴν ἀμφοῖν ἀνάλογον ὁμοιότητα” (“metaphor is the transfer [of a word] from its proper meaning to something else, according to the principle of analogy or similarity between the two” Ps.-Plutarch, The Life of Homer 19).

13 Cf. “οὐ πόρρωθεν δεῖ ἀλλ᾽ ἐκ τῶν συγγενῶν καὶ τῶν ὁµοειδῶν µεταφέρειν” (“it is necessary to transfer words that are not far-fetched [literally, ‘brought from afar’], but rather taken from related and similar things” Aristotle, Rhetoric 1405a34); cf. “µήτε µὴν πόρρωθεν µετενηνεγµέναις, ἀλλ᾽αὐτόθεν καὶ ἐκ τοῦ ὁ µοίου...” (“nor should they [metaphors] be far-fetched but taken from the same general area according to the principle of analogy” Demetrius, On Style 78).

14 OLD s.v. “translatio” 4: “transferred or figurative use (of a word)”; “translaticius” 1: “(of words) Transferred in sense”.

15 This possibly engages with the old metaphor that equates words and coins; as noted, the word “ualere” (“to be worth”, “to mean”) was used of both, as was “δύνασθαι”.

16 OLD s.v. “mutuor” b.

17 Cf. Festus 136.23-25 Lindsay: “‘metaphoram’ quam Graeci uocant, nos ‘tralationem’, id est domo mutuatum uerbum...” (“that which the Greeks term ‘metaphor’ we call ‘transference’, that is, a word borrowed from its home...”).

18 See Mankin (2011), 247.

19 See Quintilian, Institutio Oratoria 8.6.11: “praecipueque ex his oritur mira sublimitas, quae audaci et proxime periculum tralatione tolluntur, cum rebus sensu carentibus actum quendam et animos damus, qualis est ‘pontem indignatus Araxes’...” (“and above all a wondrous sublimity is effected when things are lifted by an audacious and frankly dangerous transference, for example when we give objects that lack sensation agency and emotion, such as ‘the Araxes, scornful of being bridged’”; cf. Lausberg (2008), § 559.

20 Cited byInnes (2003), 7.

21 Cf. “etenim uerecunda debet esse translatio, ut deducta esse in alienum locum, non inrupisse, atque ut precario, non ui, uenisse uideatur” (“indeed, a metaphor ought to be modest: it should appear to have been led into a foreign place rather than to have forced its way in, and to have come by entreaty and not by force” Cicero, De Oratore 3.165). This type of description goes back to Theophrastus; cf. Philodemus, On Rhetoric 173 Sudhaus = Theophrastus, fr. 689A Fortenbaugh, where the peripatetic philosopher likewise describes metaphors as “visiting a house [of someone else]”. For a different Greek example of personification, cf. “...ἀλλὰ µεµένηκεν ἡ µεταφορὰ κατέχουσα τὸν τοῦ κυρίου τόπον, ὡς ‘ὁ τῆς ἀµπέλου ὀφθαλµός’” (“...but the metaphor becomes established, holding the place of the proper term, for example ‘the eye of the vine’” Demetrius, On Style 87).

22 See Seneca, Epistulae 59.6; 114.10. On metaphor in Seneca, see Armisen-Marchetti (1989), (1991), (2015); Bartsch (2009).

23 Cf. Seneca, Epistulae 58.1: “mille res inciderunt, cum forte de Platone loqueremur, quae nomina desiderarent nec haberent...” (“when we happened to be talking about Plato, a thousand things came up which desired names but did not have them”).

24 Cf. “τοῖς δ᾽ ὄνοµα ἄνθρωποι κατέθεντ᾽ ἐπίσηµον ἑκάστῳ” (“and for these things, human beings have applied/imposed a name that marks/distinguishes each one” Parmenides 28 B19.3 DK); cf. Parmenides 28 B8.38-41 DK. The locution is also found in the Derveni Papyrus: “‘χρῆσαι’ δὲ καὶ ‘ἀρκέσαι’ ταὐτὸ[δύ]ναται. σκέψασθαι δὲ χρὴ ἐφ᾽ ὧι κεῖτα[ι τὸ] ‘ἀρκέσαι’ καὶ τὸ ‘χρῆσαι’” (“now, ‘to furnish’ and ‘to supply’ [oracles] mean the same thing. But it is necessary to examine the thing to which ‘to furnish’ and ‘to supply’ apply [i.e. on which they lie] Derveni Papyrus 11.5-7 Betegh).

25 Seneca also mentions the role of metaphor in setting something before the eyes; for similar statements about metaphor’s usefulness “demonstrandae rei causa”, see De Beneficiis 4.12.1 and Epistulae 59.6.

26 Cf. “...quibus etiam uerba parienda sunt inponendaque noua rebus nouis nomina” (“...by whom words must be issued and new names are to be set on top of new things” Cicero, De Finibus 3.3).

27 Compare Hermogenes’ usage of the verb “τιθέναι” (“to set”) for the “placement” of a word: “τροπὴ δέ ἐστι τὸ µὴ ἐξ ὑποκειµένου πράγµατος ἀλλοτρίου δὲ σηµαντικὸν ὄνοµα θεῖναι... ὃ καλεῖται καὶ ‘µεταφορὰ’ παρὰ τοῖς γραµµατικοῖς” (“a trope is a matter of placing a word that means something else and is not derived from the subject at hand... this is called ‘metaphor’ by the rhetoricians” On Invention 4.10.1-4).

28 Compare the vocabulary of the following passage: “δεῖ δὲ καὶ τὰ ἐπίθετα καὶ τὰς µεταφορὰς ἁρµοττούσας λέγειν” (“but we need to use epithets and metaphors that are fitting” Aristotle, Rhetoric 1405a10-11). Standard translations for “ἁρµόζειν” are “to fit together”, “to set”, and so on; cf. LSJ s.v. “ἁρµόζω”.

29 LSJ s.v. “σύνθετος” 2: “compound”; cf. “πρῶτον µὲν οὖν τρόποι ὀνοµάτων εἰσὶ τρεῖς, ἁπλοῦς ἢ σύνθετος ἢ µεταφέρων” (“for a start, there are three types of word – the simple, compound, and metaphorical” Ps.-Aristotle, Rhetoric to Alexander 1434b33-34). Cf. “τὸ ἐπίθετον” (“epithet”, “adjective”); LSJ s.v. “ἐπίθετος” III.

30 See Lausberg (2008), §562; cf. Stanford (1936), 37-39; Barwick (1957), 90.

31 Cf. “τὸ γὰρ ‘µυρίον’ ‘πολύ’ ἐστιν, ᾧ νῦν ἀντὶ τοῦ ‘πολλοῦ’ κέχρηται” (“and the poet has used [the term ‘µυρίον’] here instead of ‘πολύ’, for ‘µυρίον’ means ‘πολύ’” Aristotle, Poetics 1457b12-13); for a use in Plato (of the employment of a Mytilenaean word), see Protagoras 346d.

32 Cicero could simply employ the term “translatio” (“metaphor”) and its relatives for both operations: “tralata dico, ut saepe iam, quae per similitudinem ab alia re aut suauitatis aut inopiae causa transferuntur” (“by ‘transferred’ I mean now, as often before, words transferred by resemblance from another thing inorder to produce a pleasing effect, or because of lack of a proper word” Orator 92).

33 Bettini (2012), 35. The other Greek terms for “to translate” were “ἑρµηνεύειν” and “µεθερµηνεύειν”. For the Latin vocabulary, see McElduff (2013), 189-196.

34 The use of “µεταφέρειν” in the sense of “to change” was an old one; see, for example, Sophocles, Philoctetes 961-962: “πρὶν µάθοιµ᾽ εἰκαὶ πάλιν γνώµην µετοίσεις” (“not before I can learn if you will change your mind”); cf. LSJ s.v. “µεταφέρω” 2: “change, alter”.

35 The relevant Romance verbs stem not from “transferre” but from “traducere” (“to lead across”). For some thoughts on this, see Bettini (2012), vii-ix.

36 On this, see Derrida (1974), 48-49; Lloyd (1996), 205-222.

37 See Barwick (1957). On Aristotle’s role in creating the categories “natural” and “deviant”, see Lloyd (1990), 20-22.

38 Lausberg (2008), §533. Compare also the terms “ἰδιωτικός” (“non-specialized”) and “communis” (“shared”) in rhetorical language.

39 LSJ s.v. “ἔτυµος” II (“τὸ ἔτυµον”): “the true sense of a word according to its origin, its etymology”.

40 That is, Isis and Osiris were named “ἀπό τινος ἐτύµου τεθείσης ἑκατέρας τῆς προσηγορίας ταύτης” (“...on the basis of the true meaning of each name” Diodorus Siculus, Library 1.11.1).

41 Cf. DELL s.v. “κύριος”. Plato uses the term of the name giver himself at Cratylus 389d in the sense of “authoritative”.

42 Lausberg (2008), §1236-1237; Cope & Sandys (1970), 3.18. Aristotle, Poetics 1457b3-4, defines “κύριον” as denoting a word that everyone uses.

43 Concerning “ξενικός”, cf. “ξενικὸν δὲ λέγω γλῶτταν καὶ µεταφορὰν καὶ ἐπέκτασιν καὶ πᾶν τὸ παρὰ τὸ κύριον” (“by ‘foreign/strange’, I mean loan-words, metaphor, lengthenings, and everything that goes against standard usage” Aristotle, Poetics 1458a22-23); “καὶ τὸ σαφὲς καὶ τὸ ἡδὺ καὶ τὸ ξενικὸν ἔχει µάλιστα ἡ µεταφορά” (“metaphor in particular promotes clarity, sweetness, and foreignness” Aristotle, Rhetoric 1405a8-9).

44 LSJ s.v. “κυριότης”.

45 LSJ s.v. “κυριολεξία’; “κυριολογία”.

46 Cf. “κυριολογία µὲν οὖν ἐστιν ἡ διὰ τῆς πρώτης θέσεως τῶν ὀνοµάτων τὰ πράγµατα σηµαίνουσα...” (“the use of literal expressions expresses their objects according to the original imposition of names” Tryphon, On Tropes 191.6-7 Spengel).

47 Cf. “τὸ δὲ κύριον καὶ τὸ οἰκεῖον καὶ µεταφορὰ µόνα χρήσιµα πρὸς τὴν τῶν ψιλῶν λόγων λέξιν” (“standard and native terms, as well as metaphor, are alone to be employed in the diction of prose” Aristotle, Rhetoric 1404b31-33).

48 LSJ s.v. “οἰκεῖος” IV.3: “a word in its proper, literal sense, opp. metaphor”; cf. Cope & Sandys (1970), 3.18.

49 LSJ s.v. “οἰκειότης” II: “of words and phrases, proper sense”.

50 As Cope & Sandys (1970), 3.19 point out, Roman authors could translate both “κύρια” and “οἰκεῖα” as “propria”.

51 OLD s.v. “proprius” 8. The OLD provides related usages: “quae sunt hominum propria... scribam” (“I shall describe... the terms that apply specifically to human beings” Varro, De Lingua Latina 5.80); “dicitur terra proprio nomine, quae nullo alio uocabulo neque cognomine adiecto appellatur” (“the earth, which is named by no other word or description, is described by its proper name” Varro, De Re Rustica 1.9.1).

52 OLD s.v. “proprietas” 2. For the adverb “proprie”, compare the following: “tertium genus est eorum, quae proprie sunt” (“the third class consists of those things that exist in the proper sense of the term” Seneca, Epistulae 58.18).

53 OLD s.v. “proprietas” 1.

54 Cf. “nam si res suum nomen et uocabulum proprium non habet…” (“for if a thing does not have its own name and a proper term…” Cicero, De Oratore 3.159); “res ut omnes suis certis ac propriis uocabulis nominentur” (“so that all things be named by their own definite and proper words” Cicero, Pro Caecina 51).

55 For this narrative, see the following paragraph and Appendix III.

56 Cf. Mankin (2011), 235. The same thing can be noted in Varro: there are three types of words – “unum uernaculum ac domi natum, alterum aduenticium, tertium nothum ex peregrino hic natum” (“one native and born here at home, one immigrant, and a third born out of wedlock from a foreign father” Varro, De Lingua Latina 10.69).

57 See Appendix III. The main texts for this aspect of Stoic thought on this topic are Origen, Contra Celsum 1.24, Augustine, De Dialectica 6, and Varro, De Lingua Latina. For the narrative, see Barwick (1957) and Blank (1982), 22-23.

58 There is an irony in Quintilian’s opening statement on the relationship between lucidity and propriety: “perspicuitas inuerbis praecipuam habet proprietatem, sed proprietas ipsa non simpliciter accipitur” (“one prime characteristic of lucidity when it comes to words is propriety, yet the word ‘propriety’ itself does not have a single meaning” Quintilian, Institutio Oratoria 8.2.1).

59 On Demetrius’ discussion of metaphor, see Schenkeveld (1964), 88-99.

60 Cf. “etiam quae bene translata sunt, propria dici solent” (“those words too that are transferred well are often regarded as proper” Quintilian, Institutio Oratoria 8.2.11).

61 Note that throughout this description, Demetrius makes συνήθεια itself an active agent (“µεταφέρουσα”, “λανθάνει”, “λέγουσα”).

62 Schenkeveld (1964), 97, argues that Demetrius’ point is that “the κύρια, the literal, usual, proper expressions had existed, but had been definitely replaced by metaphors and not that there existed no proper expressions for e.g. the ‘eye of the vine’ and that common speech had invented names for these ἀνώνυµα by creating metaphors”; see, however, Innes (1988), 315.

63 My thanks to John Schafer for pointing this excerpt out to me. On this passage, see Bartsch (2009), 192. At De Beneficiis 4.12.1, Seneca dwells on the metaphorical usage of a term (“creditum”) “for the sake of making something clear”. On this, see Henderson (2004), 151-153.

64 See Lloyd (1996), 213: “Aristotle does not explicitly draw attention to the fact that his whole account of metaphora in general, and of its advantages and dangers in use, itself draws heavily on metaphorai. But he cannot have been unaware of the fact, and we must now ask why”.

65 Cf. Aristotle, Posterior Analytics 97b37-39: “εἰ δὲ µὴ διαλέγεσθαι δεῖ µεταφοραῖς, δῆλον ὅτι οὐδ᾽ ὁρίζεσθαι οὔτε µεταφοραῖς οὔτε ὅσα λέγεται µεταφοραῖς. διαλέγεσθαι γὰρ ἀνάγκη ἔσται µεταφοραῖς” (“and when it is not necessary to argue in metaphors, it is clear that one should not define by metaphors or what is said in metaphors; for then it will be necessary to argue in metaphors”); for similar criticism of metaphor in scientific writing, cf. Aristotle, Topics 139b32-140a2, on the need for clarity (“every metaphorical expression is obscure”); Meteorology 357a26-28, against Empedocles (“perhaps to say that is to speak adequately for poetic purposes – for metaphor is poetic – but it is not adequate for understand the nature [of the thing]”). On this, see Lloyd (1990), 21-22.

66 Lloyd (1996), 215.

67 Lloyd (1996), 217: “The encouragement to see the work as self-referential is minimal, even though he does end with a peroration in his discussion of peroration (‘I have spoken, you have heard, you have the case, decide’ 1420b3-4)”.

68 Lloyd (1996), 219-222. The relevant excerpt from Dexippus can be found at the beginning of Chapter 5.

69 Cf. OLD s. v. “quasi” 9: “(used to tone down or apologize for a word or phr. used in an unusual way, esp. in a simile or metaphor) In a manner of speaking, as it were”.

70 Mankin (2011), 235, takes “quasi” here with the entire clause; the term “uocabula” hardly needs qualification but “paene una nata cum rebus ipsis” certainly does.

71 On this absence, see Innes (2003).

72 Brink (1971), 285-286. Cf. Rudd (1989), 189: “By using dominantia instead of the usual propria, H. has both enunciated and illustrated the principle of not confining oneself to plain words”.

73 For similar metaphors in Horace, replacing theory with practice, see Innes (2003), 21-27: “the theory of metaphor seems moribund, and Horace may omit reference to that theory as one which is both too stale and hackneyed for innovative poetic treatment” (27).

74 For an introduction to modern theories of metaphor, see Pender (2000), 3-27.

75 Richards (1936/2001), 6-7.

76 For a lucid account, see Ricoeur (1977/2006), 89; cf. Silk (1974), 6-14.

77 This is also the view of Stanford (1936), 8, who cites Augustine, De Dialectica 9.

78 Richards (1936/2001), 63: “Thought is metaphoric, and proceeds by comparison, and the metaphors of language derive therefrom. To improve the theory of metaphor we must remember this”. Ricoeur (1977/2006), 92, sums up the point: “Thus, we are not dealing any longer with a simple transfer of words, but with a commerce between thoughts, that is, a transaction between contexts”.

79 Richards (1936/2001), 64.

80 Richards (1936/2001), 64.

81 Ricoeur (1977/2006), 398 n. 27.

82 Cf. Ricoeur (1977/2006), 93. See, however, Silk (1974), 12.

83 Silk (1974), 13, argues for the usefulness of the terms.

84 “To speak of the ‘interaction’ of two thoughts ‘active together’ (or, again, of their ‘interillumination’ or ‘co-operation’) is to use a metaphor emphasizing the dynamic aspects of a good reader's response to a nontrivial metaphor. I have no quarrel with the use of metaphors (if they are good ones) in talking about metaphor. But it may be as well to use several, lest we are misled by the adventitious charms of our favourities”; Black (1954-1955), 286.

85 “Let us try, for instance, to think of a metaphor as a filter”; Black (1954-1955), 286.

86 Black (1954-1955), 288.

87 Black (1954-1955), 276.

88 This terminology is used by Lakoff & Johnson (1980), for example.

89 For example, Aristotle’s example of Gorgias’ cry “Fie Philomela!” when a nightingale excreted on his head; birds cannot feel shame, and Philomela did not excrete on Gorgias’ head, but the “blend” nevertheless makes sense; see Fauconnier & Turner (2003), 36-37.

90 Fauconnier & Turner (2003), 40-44.

Table des illustrations

URL http://books.openedition.org/chbeck/docannexe/image/1622/img-1.jpg
Fichier image/jpeg, 20k

© C.H.Beck, 2016

Conditions d’utilisation : http://www.openedition.org/6540

Acheter

Volume papier

amazon.fr