Translation and the promise of analogy

ABSTRACT

Scholars of literature tend to see analogy as tied to its counterpart, metaphor, a literary device used to enrich poetic expression. Scholars in Translation Studies, on the other hand, tend to overlook analogy and instead align translation with metaphor, itself currently a controversial notion in the discipline, which, moreover, also operates on the same one-to-one logic as its counterpart, equivalence. From both points of view, analogy seems interesting, but unworthy of further critical investigation. However, as this paper argues, analogy holds promise for the discipline of Translation Studies and its pedagogy, first, because analogy was designed from the time of Aristotle to be inventive, and second, because analogy was understood by Aristotle’s medieval commentators as endowed with an intermediary nature associated with ambiguity and polysemy. Looking to analogy may allow us to respond more creatively and flexibility to the ever-changing nature of meaning today, not only in the theory and the practice of translation, but also in how we teach both.

KEYWORDS:

• translation
• analogy
• metaphor
• binaries
• equivalence

However, as this paper argues, analogy holds promise for the discipline of Translation Studies and its pedagogy, first, because analogy was designed from the time of Aristotle to be inventive, and second, because analogy was understood by Aristotle’s medieval commentators as endowed with an intermediary nature associated with ambiguity and polysemy. Looking to analogy may allow us to respond more creatively and flexibility to the ever-changing nature of meaning today, not only in the theory and the practice of translation, but also in how we teach both.

This paper aims to build a case for the notion of analogy as a conceptual touchstone for translation studies. Despite its long history in studies of language and semantics and its importance for the sciences, philosophy, and rhetoric, analogy is infrequently mentioned by practitioners and theorists of translation alike, either overlooked entirely (as in Guldin 2019) or subsumed under metaphor (as in Venuti 2019, 17). The general disregard for analogy may stem from its conflation with metaphor, a conflation which would have been impossible in the past; it also may derive from the comfort translators drawn from a conceptual logic based on correspondence, a logic that underlies not only metaphor but also equivalence.

This study does not argue that we should jettison the association of translation with metaphor, nor that metaphor and equivalence have no place in translation, and for the latter, especially in its pedagogy. Both metaphor and equivalence are far too important for translation to be utterly abolished. For one, we see more clearly the metaphoric level of language when we are faced with translating a foreign language. As Jean Paulhan wrote in 1938, whether it is a matter of translating the Kikuyu of the Bantu tribes or Cherokee into modern French, the slang of butchers and thieves into literary French, or sixteenth-century French into twentieth-century French, we are faced with an unfamiliar language that appears more imagistic, more mystical, and thus more metaphoric than our own (Paulhan 2009, 260–261; see also Blanchot 1949). Similarly, we know from our teaching the importance of the concept of equivalence, as it may provide practical solutions to the translator confronted with the differences, semantic and cultural, of a foreign language, even if it is used chiefly “because most translators are used to it” (Baker 2018, 5).

Analogy is nevertheless important for translation, and primarily because it offers nuance. Let me clarify. Translation has, I think, one problem with two faces: metaphor and equivalence. We know that scholars have criticized metaphor for making translation Eurocentric (Tymoczko 2007, 129–131); for unmooring the translator from ideology, and thus making of the translator an in-between figure (Tymoczko 2003; Baker 2005); and for instrumentalizing translation (Venuti 2019). We may also chafe under the constraints imposed by choosing a particular metaphor to describe translation and wonder how we may move beyond the restrictions that this metaphor places on translation (Sakai 2018, 3). Furthermore, equivalence is not only a central but controversial notion in translation (Kenny 2009, 96) but is also one that relies upon the same internal logic as metaphor.

Considering analogy as having its own discrete nature different from that of metaphor may allow us to move beyond the various pitfalls posed by the notion of translation as metaphor. Analogy may also help free us, and particularly in our teaching, from the imperatives of what will be described below as metaphor’s soul mate: equivalence. That is, metaphor as the hallmark quality of translation is very much tied to the imperative of equivalence in translation with which we find ourselves wrangling so often in our teaching practice. I suggest here, then, that analogy could be a significant epistemological tool for translation, if we look to the past and to the important features attributed to analogy in Antiquity and in the Middle Ages. Grounded in a diachronic reading, this study examines the development of analogy as a concept and, more importantly, as an epistemological tool.

In order to appreciate fully the promise of analogy for translation, we must begin with what might be termed the problem of translation, that is, its association with metaphor and its reliance upon equivalence. If the bonds between translation and metaphor were established long ago, they also prepare the ground for equivalence, the construction of which replicates that of metaphor.

1. The problem of translation

1.1 Translation and metaphor

Translation has been frequently described in metaphorical terms, as recent work devoted to the topic demonstrate (see especially St. André 2010; see also Guldin 2010, 2015). We find translation as bridge-building (Hönig 1997); as cannibalism (Vieira 1999), as filter (Sakai 2018), as landscape (Kershaw and Saldanha 2013); as simulacrum (Johnston 1999); as smuggling (Tyulenev 2010); and even as transubstantiation (Hermans 2007). Even at the conference in which this study was initially presented, the 4th International Summit of Writers, Translators, and Critics: The Collaboration among Writers, Translators and Critics: An Intercultural Perspective,” held in Nanning, China in 2019, appeared the metaphor of translation as a bridge, a gateway, a tightrope and the translator as ambassador. Perhaps the most layered, and even comedic, metaphor of that conference was that of translation as “duelo y quebrantos,” that is, as grief and sorrows, or scrambled eggs with chorizo.

Although we may feel a sense of incomprehension (how can one specific metaphor in question serve as the metaphor for translation? [see Sakai 2018, 3]), using metaphor to describe translation is hardly unique to the modern. When we turn to historical accounts, we find Cicero portraying word-for-word translation as a counting out of coins, unlike paying by weight, which is likened to sense-for-sense translation (Cicero 1949, 365). St. Jerome, the so-called patron saint of translation, imagined word-for-word translation as concealing meaning just as an overabundant pasture strangles the crops (Jerome 2012, 24). For Leonardo Bruni, a translation is akin to the copy of a painting; as a result, a good translation expresses “the shape, attitude and stance of [the original author’s] speech, and all his lines and colours” (Bruni 2002, 59). For Dryden, on the other hand, translation is an arduous exercise, “like dancing on Ropes with fetter’d leggs” (Dryden 2012, 39); a good translation varies only the dress of the source text (Dryden 2012, 41).

Even etymology points us to the tie of translation to metaphor as translatio. Recall how the terms for metaphor and translation in Greek and in Latin (metapherein and transferre, via the participle, translatum) are, as Douglas Robinson reminded us quite some time ago, cognates (Robinson 1991, 135). And even if the Romans did not identify translatio as applied translation and translatio as a metaphorical transfer, as Copeland (1991, 235 n74) has argued, they nonetheless placed the distinct meanings of “translatio” in proximity (Bolduc 2020, 22–28).

Take an example from Antiquity: Cicero’s Crassus, who in his discussion of metaphor in De oratore 3.157–158 provides the audience with citations that turn on the understanding of metaphor (translatio) as both the transfer of ideas and the translation of linguistic units (see also Innes 1988, 317).

Sed ea transferri oportet quae aut clariorem faciunt rem, ut illa: “Inhorrescit mare, … ”—omnia fere quo essent clariora translatis per similitudinem verbis dicta sunt—, aut quo significatur magis res tota sive facti alicuius sive consilii, ut ille qui occultantem consulto ne id quod ageretur intellegi posset duobus translatis verbis similitudine ipsa indicat: “Quandoquidem is se circumvestit dictis sepit se dolo”

[But only such metaphors should be used as either make the meaning clearer, as for instance the following: “A shivering takes the sea, … ”—to make them clearer almost all the details are expressed by metaphors based upon resemblance—, or such as better convey the whole meaning of the matter, whether it consists in an action or a thought, like the man in the play who by means of two words used metaphorically indicates by mere resemblance a person purposely using concealment in order to make it impossible to understand what was going on: “Since he employs a cloak of words, a fence/Of guilefulness”] (Cicero 1942, 122–125)

In the above passage, we find metaphoric transfer in his use of the vivid metaphor in the example that begins with the invocation of the sea shivering [“inhorrescit mare”], which may be a Latin, and more precisely Roman, equivalent for Homer’s “the battle shivered,” drawn from Illiad 13.339. On the other hand, the passage concludes with what we might more properly term translation as interlingual transfer: the citation “Quandoquidem is se circumvestit dictis sepit se dolo” De oratore 3.158 is a Latin translation of Sophocles' Greek Antigone 241. Translatio as metaphoric transfer were even more closely intertwined in the Middle Ages, and especially for late classical and medieval philosophers and theologians (Ashworth 2011, 316–319), when translatio was at times understood concretely, referring to the transfer of holy relics from one church to another known as translatio sanctorum, and at others, metaphorically, as a transfer of meaning. This latter sense is best illustrated by the medieval topos of the translatio imperii et studii, which describes both the transfer of the power of the Roman Empire to the empire of Charlemagne and the transfer of the classical learning of Athens and Rome to Paris (Curtius 1990, 29–30; Campbell 2018). If translatio studii and translatio imperii converged first with Charlemagne and the Carolingian Empire, they were nonetheless frequently distinct (Stahuljak 2005, 143–145); indeed, translatio imperii, which refers to the transfer of the cultural and legal authority implicit in political power, served to explain how the succession of empires was in accordance with divine order (Depreux et al. 2010). Translatio studii, on the other hand, referred to the transfer of textual and literary knowledge and authority (Gilson 1932; Lusignan 1989; Jeauneau 1995) and originated in the story of Alcuin’s transfer of his library from York cathedral to the monastery of St Martin of Tours (Dümmler 1895, 176–178).

As foreign as it may seem to modern readers, the medieval notion of translatio reveals the extent to which our contemporary notion of translation in metaphoric terms follows a very well-established Western tradition. Nonetheless, while it may be “difficult to draw narrow conceptual limits around the word ‘translation,’ which is almost always used metaphorically” (Sakai 2018, 1), translatio also allows us to critically reassess notions that are central to contemporary metaphors of translation (Stahuljak 2004). Chief among these, and especially in our practice of teaching, is that of equivalence, for which metaphor serves as a logical a priori. That is, whereas “[t]ranslation can be a metaphor for metaphor, and conversely, metaphor a metaphor for translation” (Guldin 2015, 2), equivalence similarly defines translation, just as translation defines equivalence (Pym 1992, 37).

1.2 Translation and equivalenc

Readers may wonder how, from the seemingly fanciful accounts of medieval translatio as metaphoric transfer above, we may arrive at equivalence, which went hand-in-hand with the founding of the discipline of Translation Studies. Equivalence may seem to be a relic of the mid-twentieth century, ostensibly scientific (tied to discourses such as we might find in Eugene Nida’s 1964/2012 Toward a Science of Translating), and pertaining mostly to a linguistic approach to translation (as Roman Jakobson’s characterization [Jakobson 1959/2012, 127] of equivalence as “the cardinal problem of language and the pivotal concern of linguistics”). Equivalence may even seem irrelevant now (Snell-Hornby 1988). As Translation Studies has moved away from linguistic equivalence towards interdisciplinarity and transdisciplinarity, equivalence has “attracted limited scholarly attention in recent years,” thereby justifying its recent absence – a “sacrifice” —in the preeminent reference work in the field, Baker and Saldanha’s Routledge Encyclopedia of Translation Studies (Baker and Saldanha 2019; xxv; versus Palumbo 2009, 42–44).

The distance between theory (the theoretical disregard for equivalence) and practice is manifest in our classrooms, where equivalence continues to be a key concept. That is, this notion, foundational and even central for the discipline, and yet very much contested in the theory of translation (see Kenny 2009, 96), remains ever-present in our pedagogical practice.

Consider the texts that appear frequently on syllabi of advanced level courses. Jean-Paul Vinay and Jean Darbelnet’s iconic 1995 Comparative stylistics of French and English: A Methodology for Translation describes equivalence as one of the procedures for translating. Anthony Pym’s Exploring Translation Theory (Pym 2014), while problematizing “natural” equivalence, proposes a model that employs “directional” equivalence, particularly for localization. Baker’s (2018) In Other Words: A Coursebook on Translation is a how-to manual for the practice of translating non-literary texts and predicated on various forms of equivalence (lexical, grammatical, textual, pragmatic). Even Lawrence Venuti’s Translation Studies Reader turns on equivalence, as its very organization – chronological as well as thematic – interrogates notions of equivalence in the history of translation theory (Venuti 2012, 5). I acknowledge that at times, the foundational theories of translation have attempted to blur this one-to-one dimension of equivalence by attempting to find nuance within a paradigm of duality. Eugene Nida, for example, divides equivalence into formal, which “focuses attention on the message itself, in both form and content,” and dynamic equivalence, which is based upon “the principle of equivalent effect” (Nida 1964/2012, 144). Peter Newmark expands on Nida’s definitions, but changes the terms to semantic equivalence, focused on meaning, and communicative equivalence, focused on effect (Newmark 1981, 1988). Then, we have House (1997), who in her description of equivalence taking place relative to function, has described equivalence as covert vs. overt, covert equivalence occurring when the TT as having same function as ST (but here any cultural differences have been attenuated).

Nevertheless, although we know that it is controversial, equivalence seems a “necessary condition for translation” (Kenny 2009, 96) and endures in our classrooms. It is taught as a technique that “replicates the same situation as in the original, whilst using completely different wording” (Vinay and Darbelnet 1995, 342). Seeking an equivalent effect on the target culture audience as on the source culture audience (Nida 1964/2012, 159; Catford 1965, 50; Catford 1994, 4739) remains one of the most frequently expressed goals of our students when they reflect on their translations. And finally, equivalence continues to be “a useful category for describing translations” (Kenny 2009, 96), which means that we continue to use it to demonstrate the quality of the target text. Although “an obstacle to progress in translation studies” (Kenny 2009, 96), equivalence lingers, then, in our teaching, whether it is a matter of the concrete strategies for translating or the criteria for evaluating the quality of the translations the students produce. Clearly, as Anthony Pym has argued (2014), equivalence, if not perfect, can always assumed in translation, and I would add, especially in the practice of our teaching.

But how could equivalence have anything to do with metaphor, we might ask? What is the relationship between equivalence and metaphor?

1.3 Metaphor and equivalence

Traces of metaphor in relation to equivalence are easy to find, appearing as early as the founding of Translation Studies as a discipline in 1972, when James Holmes described equivalence as “translation matching” (Holmes 1988, 76), as if translation were a game of association. Even critics of equivalence resort to metaphor in order to describe it. Consider Ben Conisbee Baer, who advances that “the inescapable desire and wager of translation – its fantasy – is for an absolute transparency of meaning, the impossible production of equivalence without residue” (Baer 2014, 239). And recall Hélène Cixous and Jacques Derrida’s assertion that translation “bets on a received truth, a truth that is stabilized, firm and reliable [bebaios], the truth of a meaning that, unscathed and immune, would be transmitted from one so-called language to another in general, with no veil interposed, without anything sticking or being erased, and resisting the passage” (Cixous and Derrida 2001, 55). In both quotations, equivalence in translation is likened to a form of gambling, in which meaning alternates between resembling (non)transparency and (ill-)health. As a result, if “[w]e cannot forget that the term translation contains a doubled sense of meaning” (Sakai 2018, 1), it is because the disciplinary project of translation (including its focus on equivalence) was designed, unwittingly perhaps, on the ideas of resemblance (or comparison) and substitution that are the hallmarks of metaphor.

Seeing how even equivalence is expressed in terms of metaphor demonstrates once again the inseparability, etymologically and ideologically, of translation and metaphor (Cheyfitz 1991), but this does not clarify precisely how metaphor and equivalence operate, nor the underlying principles. In response, it would be easy to refer to Douglas Robinson’s assertion, made long ago, that “metaphor is the supertrope driving the Western impulse toward translational equivalence: the attempt to bring two radically different texts, written in two different times and place, in two different languages, by two different people for two different cultures, into a mutually defining relationship” (Robinson 1991, 135). Such an assertion is powerful, and helps to elucidate how equivalence is tied to metaphor, but it does not explain the precise logic underlying this association.

That logic is one of binaries (see also Teplova 2009, 247), of a relationship of A:B, the structure of which is based on duality, an “instrumentalist dichotomy” (Venuti 2019, 11), which incomprehensibly and paradoxically comprises the movement over the border of this duality, straddling as much sameness as it does difference.

We find this logic in descriptions of metaphor. In the third book of his Poetics (1457b), Aristotle argues that metaphor is “applying to something a noun that properly applies to something else” (Aristotle 2020, 43). His characterization of the first three (of four) types of metaphor in his Rhetoric 3.11 (Aristotle 2018, 137–138) turns even more clearly upon one-to-one relation. That is, Aristotle describes metaphor as a type of transfer, which takes different forms: from genus to species, from species to genus, and from species to species. He goes so far as to caution in Rhetoric 3.11 (Aristotle 2018, 139) that “metaphors should be drawn from things that are familiar and not obvious. In philosophy too it takes a sharp mind to observe similarities between widely disparate things ….” Contemporary notions of metaphor duplicate this one-to-one relationship first established by Aristotle. Even if the most important writers on metaphor in the early to mid-twentieth century interrogate the traditional, Aristotelian view of metaphor, they nevertheless retain its binary aspect. I.A. Richards, for example, portrays metaphors as consisting of two halves, which he describes as tenor and vehicle (Richards 1976, 96–101); on the other hand, Max Black uses the terms of focus and frame (1954–1955, 276). Chaïm Perelman and Lucie Olbrechts-Tyteca designate this one-to-one relation as concerning the theme and the phore (Perelman and Olbrechts-Tyteca 2008, 501); in fact, Perelman divulges most explicitly that metaphor is expressed simply as “A is B” (Perelman 1969, 4).

We also find this logic in translation, where the list of binaries deriving from the imperative of equivalence and associated with translation is long, including source (text; author; culture; language)/target (text; author; culture; language); word-for-word/sense-for-sense (Cicero 1949, 365); domestication/foreignization (Schleiermacher 2012); direct translation/oblique translation (Vinay and Darbelnet 1995, 31); agency/invisibility (Venuti 1995). Although Toury interrogates the type and degree of equivalence between source text and target text (Toury 1980, 47), he nevertheless maintains the one-to-one relationship between them. Although equivalence may be negotiated by the translator (Pym 1992; Eco 2003), it still remains transactional. What appears to be clear for translation theorists about equivalence – that equivalence is at best convenient (Baker 2018, 5–6), at worst damaging (Gentzler 2001); that equivalence is a linguistic pipe dream that assumes that translations can render a meaning that is not only the same but also stable between languages and cultures (see, for example, Rabin 1958)—unveils a mechanical relationship between source text and target text. Even the term “equivalence” itself is predicated on a binary logic, as, according to the Centre National de Ressources Textuelle et Lexicales (CNRTL) database, it derives etymologically from the medieval Latin aequivalentia, meaning of equal quality or value, from the verb aequivalere “to be equal.”

Although troubled by the difficulty of “establishing relevant units of comparison, specifying a definition of sameness, and enumerating relevant qualities” (Halverson 1997, 210), equivalence remains true to its lexical origins, as a relationship of sameness or similarity (Palumbo 2009, 42). Similarly, metaphor for modern theorists is grounded in sameness. In his Models and Metaphors, Max Black explains that what is happening in these metaphors is fusion (Black 1962, 44–45); Lucie Olbrechts-Tyteca will later describe the reification of metaphor (Olbrechts-Tyteca 1974, 308). Whence Chaïm Perelman and Lucie Olbrechts-Tyteca’s proposal, inverting Aristotle’s paradigm, that metaphor is a condensed analogy, which has become a self-evident fact (or donnée) (Perelman and Olbrechts-Tyteca 2008, 535–536). Equivalence may not be a metaphor, but clearly both operate based on binaries. Furthermore, the dualistic logic that we find in both equivalence and metaphor lingers still, as equivalence continues to haunt the pedagogy of translation, and thus remains a question in the practice of beginning translators and their training. As a result, following Derrida (1990, 379), who declared analogy as a site of transition for translation (“l’analogie, l’affinité symbolique, le lieu de passage pour une traduction”), it is time, I think, to turn to analogy.

2. Analogy

The history of the notion of analogy is not only very long but also very complicated, far more so than metaphor, especially in that it involves multiple forms: the analogy of meaning, or of names (analogia nominum), touching on language; the analogy of reasoning (analogia rationis), typically used for questions of logic, psychology, and epistemology; and finally, the analogy of being, among things (analogia entis), which touches on the ontological or metaphysical (Ross 2011, 17–18; see also Lonfat 2004). These categories are multiplied into further aspects, so that analogy becomes associated not only with proportion (analogia proportionalitatis) but also, in the Middle Ages, with (among other types) attribution (analogia attributionis).

At this point, readers from the discipline of Translation Studies might find themselves with heads spinning. Rest assured, this essay does not attempt, or even pretend to attempt, to provide a comprehensive survey of the history of analogy, which is especially complicated in the Middle Ages (see Ashworth 1991, 1992). In the spirit of a prolegomenon, in what follows I explore specific definitions of analogy from Antiquity and the Middle Ages to the modern which complicate a simple, literary understanding of the term, and which, as a result, I argue, hold promise for translation and its pedagogy.

Recent studies in cognition, especially those that highlight analogy’s importance for cognition, consider analogy as originally tied to poetry (Holyoak, Gentner, and Kokinov 2001, 5); however, historically analogy operated within multiple realms, from science and logic to philosophy and rhetoric. A diachronic exploration of analogy in these disciplines demonstrates that analogy is much more nuanced than simple literary ornament, showing it to be a tool of complexity and creativity that transcends the simple one-to-one relation of metaphor, with which it is often associated and to which it is frequently subordinated. In Antiquity and in the Middle Ages, analogy was considered as distinct from metaphor, its attributes more complex, and even more creative. As we will see below, unlike metaphor, analogy moves beyond conceptual one-to-one relationships; situated as an intermediary between concepts, and it also allows for ambiguity and fosters polysemy.

It is important from the outset to note that, as Libera has argued, the theory of analogy is not originally Aristotelian; instead, the creation of theories of analogy arose with the commentators and translations of Aristotle’s work, which occurred over time and under the influence of translation (Libera 1989, 319–320). Moreover, medieval conceptions also add to Aristotle’s (and our) notion of analogy, because, in their synthesis of Aristotle’s metaphysical/scientific and poetic/rhetorical visions of analogy, they relate analogy both to things and to words, and see in analogy an intermediary, a halfway point between univocity (the idea of sameness in meaning, that is, a term that has but one meaning) and equivocity (a term that has many meanings). Whence the focus in the present study on the development of theories of notions of analogy from Aristotle in the Middle Ages, even if what I present here makes no claim to be exhaustive on the subject.

Analogy may seem to be traditionally aligned with metaphor, if not incorporated within it. As we have seen from the above, in his Rhetoric, Aristotle subsumes analogy within metaphor, so that metaphor is either transference (again, from genus to species, from species to genus, and from species to species) or analogy. This definition that places analogy within metaphor serves as foundational for writers on metaphor in translation, including Guldin (2015, 4–5), despite the fact that Aristotle also distinguishes them, and even discusses analogy as a separate entity. As a result, to understand analogy more fully, we must begin by looking more carefully at analogy in and beyond Aristotle’s Rhetoric and Poetics.

2.1 Classical foundations of analogy: proportion (moving beyond one-to-one relations)

Aristotle attributes several defining characteristics to analogy that are significant for the argument at hand. First, it is important to recognize that analogy is in some ways more important than metaphor for Aristotle because, when considered as one of the four types of metaphor, it is the most well liked and provides vividness in oratory (Aristotle 2018, 136–138). However, and more importantly, analogy also complicates the one-to-one relation of metaphor, and refers to a similarity of two proportions involving at least four terms, in which “B is related to A as D is related to C” (Aristotle 2020, 44). We can understand analogy, then, as not only tied to similarity but also to proportion. The term itself, as Chaïm Perelman has indicated, derives from the Greek word for proportion and is often used to describe mathematical proportion (Perelman 1977, 523); as a result, it may seem that analogy qua proportion, etymologically and philosophically, is most easily related to mathematics.

Analogy as proportion was also important as a means of scientific classification. For Aristotle, analogy was not only a type of metaphor, then, but also a technique establishing the families – the genus and species with which we are familiar – of the animal kingdom. Analogy may be used to classify and define the structure and function among animals, based either on two animals having common properties or on a similarity in the relation of parts of members to the whole species (Hesse 1965, 330). For example, as Aristotle notes in the Historia animalium I.1 (486b), the scale is to the fish what a feather is to a bird (Aristotle 1965, 7).

Aristotle uses analogy not only in scientific but also in metaphysical analysis (Hesse 1965), employing analogy (again, in terms of proportion) in his description of the relationships between the ten categories of being. His understanding of analogy is broadened in the sixth century CE, when Simplicius of Cilicia (d. ca. 560CE), in his commentary on Aristotle’s categories, uses analogy to demonstrate how relatives, one of Aristotle’s ten categories of things that exist in the sensible world, may also apply to Platonic intelligibles (Simplicius 2002, 64). In other words, for Simplicius, living creatures may serve analogically as a means to understand immaterial propensities (Simplicius 2002, 110). As a result, Aristotle’s categories of being cannot be considered as synonymous (otherwise, there would only be a single genus), but the relationship between them will be established either by paronymy or by analogy (see Courtine 2005, 153–239).

[As a reminder, paronymy is a derivative term (e.g. “historian” from “history”) and, for ancient commentators of Aristotle, an intermediary between homonyms and synonyms; that is, paronymy is a relation of naming based on things having a different relation to the same thing. On the other hand, analogy is a similarity between things/terms that are heterogeneous, and is used across genera. This transference will be important in the Middle Ages, when the word and concept of analoga will come to encompass those of denominativa (naming), understood in the Middle Ages as a literal translation of παρωνυμα [paronymy] (Courtine 2005, 244).]

2.2 Classical foundations of analogy: analogy and invention

More important, analogy is inventive in a way that metaphor, held back by its tendency toward fusion, cannot be. Of this, Aristotle provides in his Poetics 1457b several lovely examples:

I mean, for instance, a wine bowl is to Dionysus what a shield is to Ares, so you may call a wine bowl “the spear of Dionysus” or a shield “the wine cup of Ares.” Or again, old age is to life as evening is to day, so you can speak of evening as the day’s old age, or, like Empedocles, call old age the evening or the twilight of life. (Aristotle 2020, 44)

These examples of analogy are not only poetic but also demonstrate the way in which analogy requires creativity. With these examples, and particularly with the second example of old age as the twilight of life, we might be reminded of Holyoak, Gentner, and Kokinov’s assertion that one of the basic functions of analogy is the transfer of emotions (Holyoak, Gentner, and Kokinov 2001, 5). Nevertheless, it is the imagination and originality necessary for the creation of these analogies that is most striking, because they transcend the one-to-one relation established by metaphor.

Similarly, we find ourselves confronted with invention within Aristotle’s use of analogy for scientific classification. That is, Aristotle suggests that analogy may be used to refer to things which did not belong to one genus and lacked a common name. Consider how he compares a fish spine, a squid’s pounce, and an animal’s bone in Book 2 of his Posterior Analytics (98a20): “There is another method of selection, viz., by analogy. It is impossible to find a single name which should be applied to pounce, spine and bones; yet the fact that these too have <common> properties implies that there is a single natural substance of this kind” (Aristotle 1960, 243). In other words, we use analogy in situations when, as in the case of a squid’s pounce, a fish’s spine, and an animal’s bone, which possess the common property of having bony parts, we cannot find one sole and identical name for this osseous nature. As Mary Hesse notes (Hesse 1965, 329), this example of analogy

seems to depend on the apparently accidental and comparatively trivial fact that there happens to be no word in the Greek language to express what is seen to be in common between the instances. The point is, however, clearly not one merely about language, for Aristotle immediately coins an appropriate phrase, namely “osseous nature,” to fill that gap. In this example he wishes rather to direct attention to the process of selecting the common properties. This is what is done “by analogy,” and it is in virtue of the analogy that a common name can be coined.

In other words, analogy at times serves to conjure a missing term; it calls forth a word, highlighting the way in which invention is once again operative in analogy. Similarly, imagination as invention will be applied to different phases of composition in the Middle Ages, so that the process of how idea in the mind will lead to the written work and in turn to its style and words will be considered as poetic analogies (Kelly 1991, 66–67). We may as a result better understand how Derrida, in his interpretation of Aristotle’s Poetics (1458a26–27), considers analogy a “‘riddle,’ a secret fecitative” [sic], in that “the essential character of [analogy] is to describe ‘a fact in an impossible combination of words’” (Derrida and Moore 1974, 44), and how Perelman and Olbrechts-Tyteca (2008, 517) will describe analogy as a tool for invention in argumentation.

2.3 Medieval developments: analogy as intermediate, ambiguous, and polysemous

Medieval glosses provide a conception of analogy that is even more nuanced and complex than Aristotle’s foundational notions. Its newfound status as intermediary and the resulting hermeneutic and translation history which aggregates additional features of ambiguity and polysemy to it expand our understanding of it so as to be significant for the present study.

First consider how analogy comes into play as an intermediate term in the Middle Ages. Whereas Aristotle introduces a “third” way in metaphysics between univocity and equivocity (Hesse 1965, 328), it is only from the time of the earliest commentaries on Aristotle in Arabic that this intermediary item between univocity and equivocity will be identified as analogy (Libera 1989, 329–334). Analogy is, then, attributed an in-between status, unlike metaphor: recall that even for Greek commentators on Aristotle, including Simplicius and Boethius, whereas an equivocal term may be related by means of the metaphoric usage of the transfer of names (translatio), they are nonetheless distinct, since translatio involves words, and equivocation, things (Ashworth 1992, 101–102).

The notion of analogy as falling between univocity and equivocity will be essential to discussions of theology in the High Middle Ages, when, analogy (per attributionem) will also be a feature of discussions of equivocation (including homonymy and polysemy), encompassing the idea of pros hen, or focal relation (Ashworth 1992, 94, 98). Whereas univocity in the Middle Ages is always in reference to God, and univocity also appears in Thomistic visions of the relationship of beauty and being (Jaroszyński 2011, 31–39), the analogy of being, especially for Aquinas, derives from the causal dependence of beings on God (Courtine 2005, 265). Clearly, analogy as intermediary, as a result of the Latin translations of Aristotle’s Metaphysics, has become in the thirteenth century associated with ens (a being or entity) rather than a quality or attribute (Ashworth 1992, 107; see also Libera Aubenque 1978, 334–336; Aubenque 1976, 198–206; Aubenque 1978).

Second, the notion of analogy as falling between univocity and equivocity is also critical for the present study, because it introduces the concomitant idea of ambiguity as located at the heart of analogy. In his commentary on Aristotle’s Topics, Alexander of Aphrodisias (ca. 200 CE) had early on distinguished a term that is neither univocal nor equivocal, and which he called “ambiguous” (Wolfson 1938, 152). It will be in the thirteenth century, however, that medieval scholars will incorporate the idea of analogy as ambiguity via Arabic translations of Aristotle. In fact, as Wolfson explains, the Arabic translation of the term for “equivocal” in Aristotle’s Categories is both as “mushtarak” and “muttafiq”; however, over time “muttafiq” loses its meaning of original of “equivocal,” acquiring instead the meaning of “ambiguous” (Wolfson 1938, 168–169). As a result, it is unsurprising that Robert Grosseteste’s commentary on Aristotle’s Posterior Analytics (ca 1220), which is the earliest extant and complete medieval Latin commentary on this work, and which was based on the vulgate translation made by James of Venice in the second quarter of the twelfth century (see Lohr 1972, 100–107; Dod 1982, 61), ties ambiguity to analogy. Grosseteste notes that Aristotle’s use of analogy to find a common term produces ambiguous names (ambiguum analogum) said according to a prior and a posterior sense. Aquinas, who will make use of such thirteenth-century translations and commentaries of Aristotle by Robert of Grosseteste and William of Moerbeke, will advance a notion of analogy that had consequently already absorbed what was called in Arabic philosophy “ambiguous” (Wolfson 1938, 171–172).

Third, that analogy is an intermediate term that is also significant for how it also produces polysemic meaning. Recall how Albert the Great (d. 1280) will describe how a term may have multiple signification through analogy or proportion in his Liber de Praedicamentis 1.2 (“Et hic quidem modus vocatur multiplex dictum secundum analogiam, sive proportionem ad unum quod principaliter in nomine significatur”; Borgnet 1890, 153a). Consider too how in his De signis Roger Bacon produces a new idea of analogy as generating knowledge and meaning. As Thomas S. Maloney indicates, when someone imposes a name for something which is naturally related to the other things (the concept of the thing, the name of the thing, the metaphysical components of the thing), this imposition unintentionally gives rise to a knowledge of a lot of things in addition to the thing originally intended to be named (Maloney 2013, 13–14). That is, in his On Signs 100–130, Bacon declares that “[o]ne should know, therefore, that when one imposition of names is made, [the name] does not signify many things univocally, because in univocation there must be one signification and one definition of the things [signified]. But when this happens there are many different significates, as is clear” (Bacon 2013, 99).

The polysemic nature of analogy as intermediary lends itself, perhaps unsurprisingly, to analogy being considered in different modes by Albert the Great (Libera 1989, 331). Similar too is the medieval three-fold division within analogous terms, which arise when a term signifies two things, the first primarily, and the second, either through attribution (in relation to the first) or via transferred meaning (Ashworth 1992, 119). This division of analogous terms into univocal, equivocal, and those that fall in between (based on a common characteristic which is found primarily [per prius] in one thing and secondarily [per posterius] in another) will become an issue of importance for medieval logicians, and in particular for Aquinas (Ashworth 1992, 119–130).

2.4 Medieval developments: complexity and the translation of “analogy”

The history of the translation of the very term “analogy” also provides further evidence of the complex and polysemic meaning it acquires in the Middle Ages. For one, the Greek word for analogy was the same as the Latin proportio, and yet the two were often conflated in translation in the Middle Ages. For example, Robert of Grosseteste’s Latin translation of Aristotle’s Nicomachean Ethics presents the definition of analogy as an equality of ratios using the term proportionalitas rather than analogy (Ashworth 1992, 99).

We cannot underestimate the importance of translation from the Arabic, which holds great influence on the understanding of analogy in the medieval West. Consider Albert the Great’s work on Porphyry’s universals, where he describes analogy as proportion that falls between equivocity and univocity as deriving from Arabic: he specifically notes that Arabs refer to analogy using the Arabic term convenientia (“Analoga autem sive proportionaliter dicta sive, ut Arabes dicunt, ‘convenientia sunt media inter univoca et aequivoca,’” Noya 2004, 11a; see also Libera 1989, 329–331). Albert’s use of convenientia may derive from Al-Ghazālī’s Latin Logica; the term convenientia, deriving from the Arabic term muttafiqa, most likely has its origin in Avicenna’s original Arabic works (Libera 1989, 332), demonstrating again the importance of translation in the history of our understanding of “analogy.”

Indeed, this co-existence of analogy with convenientia will continue in the work of such thirteenth-century scholars as Aquinas, Roger Bacon, and James of Viterbo (Libera 1989, 333–334); moreover, both Viterbo and Aquinas will expand on this association. Aquinas, who uses Moerbeke’s translation (Courtine 2005, 244), provides further facets to the definition of analogy in De principiis naturae 6, by suggesting that analogy is either proportion, comparison, or convenientia (“Aliquando enim ea quae conveniunt secundum analogiam, id est in proportione, vel comparatione vel convenientia”; Aquinas 1954, n. 367). Viterbo will add further nuance, by distinguishing a convenientia of resemblance (convenientia similitudinis) and a convenientia of attribution (convenientia attributionis) from analogy (analogiae), relation (habitudinis), and proportion (proportionis) (James of Viterbo 1983, 196; Libera 1989, 333–334).

By contrast, metaphor, particularly in the West, does not enjoy such a complicated, aggregate, and disciplinary-blurring history, and the meaning of the Latin and Greek cognates for metaphor Greek and in Latin (metapherein and translatum) remains relatively stable over time. In fact, although Arabic translations of Greek philosophy, science, and medicine may have been instrumental in developing an abstract style in Arabic literature (Walzer 1953), these translations did not produce alterations in the conception of metaphor.

3. Conclusions: the promise of analogy for translation studies

With analogy, then, we have come a long way from metaphor and equivalence. The Aristotelian vision of analogy from Aristotle and his medieval commentators is clearly a byzantine, complicated history, filled with additions to and transformations of the meaning of analogy. It is precisely this fluidity of the very meaning of analogy that may potentially offer Translation Studies a way to untangle the knot of translation as metaphor and its concomitant imperative of equivalence that still lingers in our teaching. It is easy to envisage how, because of its long and diffuse history as a concept in science, philosophy, and metaphysics, analogy encourages a disciplinary blurring which may further enrich Translation Studies’ current interdisciplinary tendencies. But what are some more specific features of analogy that hold promise for Translation Studies and its pedagogy in a way that metaphor does not?

First, analogy, unlike metaphor, was designed from the time of Aristotle to emphasize the creativity of invention. Further, in its development from the medieval commentators on Aristotle, analogy also began to take on not only a tolerance of but also a fostering of ambiguity, polysemy, and the meaning that flourishes in between meaning. As a result, analogy encourages mental leaps that take us beyond binary thinking, which is essential for translation, and especially for leaving equivalence in the (proverbial) dust. Imagine, for example, if we were to incorporate a statement on our syllabi indicating that translation is an art, and that as practitioners and scholars of translation we do not believe in the illusion that there is ever only a sole and unique translation, or a single equivalent term for another? Or if we were to incorporate within our marking criteria language indicating that we appreciate at least some degree of creativity, and, on the other hand, that we are hoping to see translations that leave behind the most surface-level, literal renderings as well as commentaries that abandon the most tired metaphors about translation and translators. Even more effective, I think, would be to make the practice of analogy concrete in our introductions to the theory and the practice. An easy way to effect this would be by providing from the outset, and regularly, multilingual examples of translation structured on analogical principles. Inviting students early on in the course to provide their own multilingual examples, perhaps based on an initial simple discussion of “bread”—Arabic khubz : Chinese miànbāo : English bread : French pain [or German brot : Italian pane, Spanish pan, etc.] (see Munday 2014, 67)—would allow us to explore how such analogies complicate a simple, singular understanding of meaning, not to mention the processes of translation. We might also evoke in further lessons how “different kinds of analogy constitute different vehicles of translation, articulate different forms of movement” (Bannet 1997, 657), recalling that Aristotle’s notion of analogy as inventive reflects a desire (also relevant for translation) for a common discourse on heterogeneous things (Aubenque 1978, 11). That analogy generates creativity, ambiguity and polysemy would thus open our pedagogy – and perhaps our marking criteria – to a breadth of possibility that is not provided by metaphor, given how it is structured according to the same logic that also underlies the imperative of equivalence.

Second, a significant marker of analogy is its ability to establish complex relations between things and concepts. Both Aristotle’s use of analogy for categorizing things in nature and Simplicius’s use of Aristotle to extend the use of analogy beyond the terrestrial suggest that analogical proportion does not reproduce metaphor’s rigid, mathematically inflected one-to-one relation. In fact, both Aristotle and Simplicius adumbrate Chaïm Perelman and Lucie Olbrechts-Tyteca’s more recent argument that analogy’s proportion is not simply related to mathematics as its etymology would suggest, but is rather based on a similarity of structural relations (Perelman and Olbrechts-Tyteca 2008, 500–501). Medieval commentaries on and translations of Aristotle will add to this the important notion of analogy as intermediary, cementing its role as an interactive agent that moves between things and concepts. They thereby emphasize its function in relationships, rather than simply as a comparison. This notion also holds potential for translation: analogy’s emphasis on a structure of relations that is multiple rather than singular or fusional suggests that translation is anything but a mechanical activity and may thereby re-establish the importance of the trained translator in a world (especially lay) too often reliant on machine translation. In this light, we may also see translation as establishing a complex chain of relations, and not only between source author and translator, or source text and translation. Rather, such a perspective also underscores how language may be multilingual, translingual, hybridic, and creole, and furthermore, how translation may be as much intermedial as intersemiotic (and more).

In addition, whereas this focus on analogy’s effect via structural relations will lead to the idea in studies of cognition that “[a]nalogy, in its most general sense, is this ability to think about relational patterns” (Holyoak, Gentner, and Kokinov 2001, 2), it also may give us further impetus to act and to choose, whether it is a matter of translating a particularly difficult text or teaching a difficult concept in Translation Studies. As Eve Tavor Bannet writes, “Analogical reasoning has everything to do with how we ‘go on’ because both in our disciplines and in our everyday uses of language, we are always using analogies to translate familiar terms, concepts, and images from one place to another place which might also be quite different – for example, from matter to man, from nature to God, from context to context, from word to image, or from image to deed” (Bannet 1997, 656). Knowing that there “may be no neutral choices in language, but there are choices” (Fahnestock 2011, 44), the translator may turn to analogy so as to overcome the paralyzing dictates of the one-to-one relation fashioned by metaphor and manifest in equivalence.

I would point out, furthermore, that modern philosophy attributes to analogy rather than metaphor a role in reasoning, as analogy allows for the value of terms to be determined and even to change, which is critical in persuasive discourse (Perelman and Olbrechts-Tyteca 2008, 512–513). That is, the speaker, in seeking an audience’s endorsement of an idea, uses – or invents – analogy to create a structural correspondence of relations that aims at, but that does not always result in, adherence and a “meeting of minds” with the audience. As a result, analogy, structured by this complicated set of incomplete, ever changing relations, may thereby locate the translation as much as the original, the translator as much as the author, the audience/readers as well other actors of the translational act, within a rhetorical situation (France 2005, 261).

Faced with the restrictive paradigm that translation as metaphor presents, analogy and analogical thinking holds, then, great promise for the discipline and for its pedagogy. Lawrence Venuti’s recent call that we stop treating translation as a metaphor (2019, x) could be met with the assertion made by Derrida (1990) that analogy is a form of translation (see Bannet 1997, 655), or even Richards’ plea that “we all live, and speak, only through our eye for resemblances” (Richards 1976, 89). Heeding Richards’ call by turning to analogy may prevent us from slipping back into the fusional dynamics that translation as metaphor/translation as equivalence poses.

Analogy may be one of those epistemological tools having the complexity and polysemic features that translation requires. Indeed, as Chantal Wright has written, “[w]e translators need to take ourselves seriously – more seriously – as intellectuals, as highly trained readers, and as artists. And I believe there are models for this out there, artists and intellectuals who are pushing the boundaries of what translation can be” (Wright 2020, 3). One such model could be, I posit, analogy, and not only because every philosophical reflection is structured, made intelligible, and expressed by means of analogies (Perelman 1969, 8, 15), nor because “without concepts there can be no thought, and without analogies there can be no concepts” (Hofstader 2013, 3), but rather, because analogy may re-endow our pedagogies of translation, and perhaps even the discipline more broadly, with the necessary creativity and flexibility to shape translation, in practice and in theory, for the ever-changing and ambiguous nature of meaning in our intercultural and intermedia world.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Michelle Bolduc

Michelle Bolduc is Professor of Translation Studies at the University of Exeter, UK. An internationally recognized scholar of translation studies and comparative medieval literature, she is also at the forefront of bringing the work of Chaïm Perelman’s and Lucie Olbrechts-Tyteca’s New Rhetoric Project into English. Author of The Medieval Poetics of Contraries (2006), she has published extensively on medieval literature and rhetoric, as well as on modern rhetoric – the New Rhetoric Project – and its translation. Her most recent book, Translation and the Rediscovery of Rhetoric (2020) presents a diachronic case study of how translation is the means by which rhetoric, as the art of reasoning, becomes a part of a lineage of—and a resource for—an ethics of civic discourse.