Aristotle and metaphor

Greek Philosophy #essays

This is the first half of an unfinished paper I was writing in an attempt to get into grad school. It’s probably not going to get finished, but I think there’s some pretty unique stuff in her about Aristotle and metaphor.

“The greatest thing by far is to have a command of metaphor. This alone cannot be imparted by another; it is the mark of genius, for to make good metaphors implies an eye for resemblances.” – Aristotle, Poetics

“The history of metaphysics, like the history of the West, is the history of these metaphors and metonymies.” –Jacques Derrida, Structure, Sign, and Play in the Discourse of the Human Sciences

There is a growing body of evidence supporting the idea that visual – rather than just literal – thinking is central to the enterprise of learning and communicating knowledge to others. While we’re quite accustomed to the idea of visual or metaphorical thinking in humanistic domains like literature and spirituality, it’s now becoming clear that similar types of visual and spatial reasoning are equally central to logical thought and communication on a number of different levels, from the extremely mundane to the scientifically revolutionary. On the level of everyday thought and communication, contemporary research in cognitive science (such as the work of George Lakoff and Mark Johnson) suggests that almost all abstract thought may actually take place by way of spatial metaphor. On the level of groundbreaking discovery, there are a number of historical examples of the way that an intuitive, imagistic grasp of a novel solution often seems to precede conceptual understanding in a direct way (and there is also some interesting theory around this phenomenon in Structural Intuition, by Martin Kemp). Johnson and Lakoff argue that our contemporary understanding of the metaphorical nature of thought calls for a complete reframing of the history of knowledge, and I agree with them to a significant extent. Based largely on their work, I  will outline a case that visual thinking and metaphor have always been fundamental to the fields of philosophy and science (among others), and thus should also be central to the way we understand and teach the history of those subjects. However, when the authors go on to suggest avenues for applying their ideas to the history of philosophy, I am less in agreement with their interpretations and methods. I will use examples from Aristotle and Plotinus to argue that Ancient Greek philosophers already had a far more nuanced understanding of metaphor than this work gives them credit for. They used metaphorical thought and language in a variety of consciously skillful ways (in some cases, even describing the functions of metaphor in terms surprisingly similar to this contemporary work and others). As Sarah Rappe says in Reading Neoplatonism, Linguistic theory was just as significant for the Neoplatonists as it is for modern philosophers,” and I suggest that the same is true of Aristotle and many other ancient thinkers. Any history of metaphorical thinking should begin with an attempt to appreciate and understand this tradition of expertise more fully.

In their book Metaphors in the Flesh, Lakoff and Johnson begin with an argument that almost all abstract thinking is actually metaphorical. Their model is that all human beings acquire basic metaphorical associations between physical experiences and abstract concepts by experiencing them simultaneously as babies. For example, one cross-cultural basic-level metaphor is that “more” is equivalent to “up.” Children share the experience that when something is added to a pile (more) the visual height of the pile appears to rise (up). Neurons in different areas of the brain related to amount and to vision or spatial processing will fire at the same time, and when this process is repeated extensively, the conceptual association becomes neurologically automatic. In the future, these children will experience the firing of neurons in both areas if, for example, they’re shown a diagram with an arrow pointing up. They may automatically assume that the conceptual meaning of such an image is that the amount of something is increasing.

In addition, these basic-level metaphors eventually combine to create a huge number of more complicated metaphors that we use on a daily basis. These complex metaphors vary somewhat by culture, but many are still very common. For example, several simpler metaphors (including similar equates to physically close and ideas can be manipulated like objects) contribute to the more complex but incredibly common metaphorical understanding that conceptual categories are like containers (so, concepts can be sorted into different categories, just like objects can be sorted into different containers).

We find it easy enough to understand that metaphors like this can be used to explain ideas like conceptual categorization to people who don’t understand them, but slightly less intuitive to notice that we actually use this kind of spatial metaphor (that is, mental mapping of abstract concepts onto our experiences of physical objects and/or actions)  all the time when we think creatively and solve problems. The container metaphor provides a particularly vivid example of this kind of metaphorical reasoning if you think about a logical problem like, “All As are Bs and all Cs are Bs. Can As be Cs?” Unless you’re so used to this kind of problem that you solve it automatically, you’ll almost certainly imagine a diagram with circles (a type of visual container) inside of other circles. Using this visual metaphor, of course, it’s easy to figure out that because circles A and C can overlap inside of circle B, we can also assume that categories A and C can overlap. Categories are not actually circles, and no metaphor is ever perfect, because no thing can ever be exactly the same as another thing it’s being compared to — for example, while it could still facilitate a number of correct intuitions, you could easily run into problems with the container metaphor and this logic question if you imagine the categories as, say, baskets rather than circles, because you would run into an area in which categories and baskets are not the same (specifically, that a concept can often be in more than one category, while an object often can’t be inside two non-overlapping baskets). Nonetheless, the container metaphor is apt enough to provide correct answers to many types of category questions, and if you use a type of container that can overlap (like a circle drawn on paper), it is even more apt. Thus, the act of reasoning metaphorically is a constant process of discovering metaphors that are apt enough to align with reality in ways that provide useful information, while also recognizing that all possible metaphors (and therefore, almost all types of conceptual understanding) have their limits.

If you think about this kind of thing often, you’ll notice that an overwhelming number of words and expressions are clearly metaphorical in origin. Even those that don’t have obvious metaphorical associations often will once you consider their etymology. For example, the word metaphor itself means to carry across. While Lakoff and Johnson describe a few abstract concepts as literal rather than metaphorical, even the example they use for this, similar, has linguistic roots that mean “close.” Can all abstract language be boiled down to ancient physical roots? It’s possible, but if so, it’s still worth noting the distinction made by theorists including Paul Ricoeur between dead and living metaphor. Over time, metaphorical language can become so common or antiquated that it doesn’t actually help us to reason anymore, in the same way that we might not think about carrying at all when we think about metaphor. On the other hand, living metaphors like containers for categories can still be used to generate new insights thousands of years after they were developed as long as we can still relate them to our physical experiences, and often even though we aren’t consciously aware that we’re using them.

This ability to reason with metaphors without being aware of them is crucial, and a major part of the argument for appreciating the history of metaphor as part of the history of knowledge. As Pierre Hadot puts it, “Philosophy is thinking in a tradition,” and while thinking in a tradition may mean adopting certain premises or explicit approaches to problem-solving, it also means participating in a culture of shared metaphors, both explicit and subconscious. Lakoff and Johnson write, “When we look into these [philosophical] questions we discover that they can’t even be formulated without first presupposing a large number of folk theories, idealized cognitive models, and metaphorically defined concepts. The same holds for the philosophical theories that arise as answers to these questions. It is our claim that philosophical theories are attempts to refine, extend, clarify, and make consistent certain common metaphors and folk theories shared within a culture.” As they go on to point out, these crucial cultural metaphors may be described in philosophical writings, but more often they may simply be assumed as a shared point of reference. Especially when implicit, a better understanding of these cultural metaphors could contribute greatly to our understanding of historical thought.

Finally, Lakoff and Johnson go on to sketch some points of departure for research into the history of metaphor in philosophy. They briefly touch on the Presocratics and Plato, but I will focus on their comments on Aristotle, for two reasons.  First, Aristotle wrote prolifically on linguistic subjects including quite a bit on analogy and metaphor, providing a crucial perspective on how ancient thinkers could have understood these subjects. Second, as a major figure in the history of logic and science, Aristotle and his work seem like an appropriate place to start an investigation on the history of metaphorical thinking as applied to this kind of extremely non-fanciful subject matter specifically.

Unfortunately, though, Lakoff and Johnson don’t seem to have a high opinion of Aristotle’s ability to understand and engage with metaphor creatively.  They write that he must have had a very different and limited concept of metaphor, as “a mere use of words, not a matter of concepts,” and specifically, “a deviant use of words, since they were applied to things they do not properly apply to.” This would have led to the idea that “if a metaphorical expression had a meaning at all, there would have to be some consistent basis for determining what the appropriate literal sense was.” Basically, they describe Aristotle’s conscious approach to metaphor as a way of substituting one word or phrase for another in a way that can be “decoded” for aesthetic effect.

Aristotle himself describes metaphor as anything but a deviant or merely decorative use of words, though. In Poetics, he lists the possible types of words as, “current, or strange, or metaphorical, or ornamental, or newly-coined, or lengthened, or contracted, or altered.” In Rhetoric, he says (in reference to the same list), “Strange words, compound words, and invented words must be used sparingly and on few occasions. […] In the language of prose, besides the regular and proper names of things, metaphorical terms only can be used with advantage.” This is because, along with literal language, they are the only type of language used in everyday speech.

Not only does Aristotle describe metaphor as a normal and desirable use of language, he mentions several reasons why it is desirable. For example, “Strange words simply puzzle us, ordinary words convey only what we know; it is from metaphor that we can best get hold of something fresh.” He describes the way a metaphor can be used to suggest the scattering of the sun’s rays by comparing this action to the sowing of seed, even though (perhaps especially because) there’s no specific word for the equivalent of “sowing” in relation to the sun. This kind of indirect description, he says, conveys liveliness; “Because the hearer expected something different, his acquisition of the new idea impresses him all the more.” These descriptions certainly hint at the application of metaphor more for its special ability to create new meaning than for its decorative effect.

Aristotle describes an example of metaphorical entailment…[more on metaphorical entailments of word choices PITF 150, examples in Rhetoric of how Aristotle understood these]

Despite all this, Johnson and Lakoff summarize their view that, “Aristotle, of course, noticed the existence of metaphoric uses of language. But given his central metaphors and the overall conceptual structure of his philosophy, he could not have come up with anything like the contemporary theory of metaphor that we have been using. Aristotle could not possibly have seen metaphor as a conceptual mapping from one conceptual domain to another, where the inferential structure of one domain is mapped onto another.”  Their reasoning for this perspective seems to be that since Aristotle regarded ideas as something “out there” in the world rather than something generated by human minds, he could not have really grasped the concept of thinking about one thing “in terms of” another; the only thoughts available would be thoughts about each individual thing, discrete in itself, and if two things could be reasoned about in similar ways, it would be because those two things were fundamentally variations on the same type thing. As they say, “For Aristotle, the father of logic, logic is the logic of the world. What we can logically think depends on the way things are in the world. Logic for Aristotle is not an abstract issue; It occurs as part of the world and has a locus in space, in time, and in objects.”

This objection to the possibility of an understanding metaphorical thought is somewhat unsatisfying. It does suggest that Aristotle might have described the nature of conceptual metaphor and its function somewhat differently than we would today (perhaps, as they say, “Aristotle could not possibly have seen metaphor as a conceptual mapping from one conceptual domain to another,” since anything something could logically be mapped to must therefore somehow be part of its proper conceptual domain), however, this doesn’t at all preclude the skillful and conscious use of metaphorical reasoning (i.e., the crucial part “where the inferential structure of one [thing] is mapped onto another”). There’s no reason that hidden connections between different things couldn’t be part of the logical pattern of the world, and an important key to how we think about them — in fact, Aristotle recognizes the logical importance of making connections between apparently disparate things, saying that, “Metaphors must be drawn, as has been said already, from things that are related to the original thing, and yet not obviously so related — just as in philosophy also an acute mind will perceive resemblances even in things far apart.” The example of a good metaphor that he gives, in this case, is a pretty straightforward form of basic spatial metaphor: that of describing states as “leveled,” comparing political equality with equal physical flatness.

While all these passages could be read as relating primarily to skillful use of language for communication and explanation, Aristotle was also well aware of the vital role of visual thinking and spatial metaphor in conceptual thinking. In De Memoria he notably wrote, “It is impossible to think without an image.”

One interesting example of how this kind of thinking can play out comes from Aristotle’s consideration of the phenomenon we now know as gravity in his Physics. Lakoff and Johnson describe it like this: “People are naturally healthy; occasionally they get sick, but they tend to become healthy again. In general, things have natural states. When they are removed from those natural states they tend to return to them. By Aristotle’s theory of metaphor as expressing similarity, states and locations are similar and so are of the same genus. […] Consequently what we know about states can be applied with certainty to locations. Since things have states they are naturally in, they must therefore have locations they are naturally in. […] A stone’s natural location is on the ground. When thrown in the air, it is removed from its natural state to which it seeks to return.” They regard this way of thinking as distinctly non-metaphorical because it starts from the premise that when two things are similar and can be reasoned about similarly, what we know about one can generally be applied to the other “with certainty,” leading to all kinds of factual errors (such as the basket error where a thing can’t be in two categories).

While we don’t have a record of Aristotle’s thoughts on the subject of whether there is a limit on the similarities between places and states, we must assume, though, that he understood that there was such a limit. There is no evidence whatsoever to suggest that, while he may have seen them as “of the same genus,” Aristotle thought that places were literally exactly the same as states. In fact, talking about them distinctly demonstrates that he understood them as two distinct categories, which must therefore differ in some way. Any error in reasoning (like so many errors in scientific reasoning, even in the present day) lies in correctly understanding the extent and nature of the similarity and the difference between the two things being compared. There is no reason to think that this type of error would be more or less likely based on whether the two things being compared are fundamentally related through some universal pattern of organization, or merely incidentally related based on apparent perceived similarities. Furthermore, Aristotle understood very well that the theories could be generated in a variety of ways, but must be checked against factual observation. As Pierre Hadot says in Philosophy as a Way of Life, “The Aristotelian researcher was no simple collector of facts. Rather, facts were amassed to make possible comparisons and analogies, to establish their causes. All this was accomplished with close collaboration between observation and reasoning — an alliance in which, says Aristotle, one must have more trust in observation of facts than in reasoning; the latter is to be trusted only insofar as it accords with the observed facts.”

In order to point out an even more remarkable example of Aristotle’s facility with metaphorical reasoning, I need to briefly backtrack and describe two details of Lakoff and Johnson’s work for comparison: their account of metaphorical duality, and their method of using knowledge of metaphorical thinking to solve one of Zeno’s paradoxes.

A metaphorical dual is a pair of metaphors describing the same concept in almost identical physical terms, except with a reversal of perspective. For example, we often imagine our passage through time as a form of spatial travel similar to a timeline. We see ourselves facing toward (and moving toward) the future. However, it’s equally possible to imagine that time is flowing toward us, and events in the future are approaching us while we “stand still.” In either case, the relative motion is the same: We and the future get closer and closer to each other; we and certain future events approach and then “pass” each other. Because these are spatial metaphors, they would mostly be different from the perspective of a specific thing/event that is neither us nor the future. Johnson and Lakoff give the example of someone who says, “Let’s move the meeting ahead a week.” This is an ambiguous expression because we sometimes imagine the movement of time in both these ways, and it’s not immediately clear whether the speaker wants to move the meeting “ahead” with respect to us, or with respect to the future events that are flowing toward us. Another common kind of metaphorical duality (besides spatial perspective) is called multiplicity/mass duality. For example, is time a flowing mass we “pass through,” or a series of individual moments we encounter one at a time?

Really, of course, it is neither kind of spatial thing — but the temptation to conflate this kind of dual can lead to a lot of confusion and ambiguity. For example, Lakoff and Johnson discuss Zeno’s paradox of the arrow, that if you imagine an arrow flying through the air at any number of points on a timeline, it will be in one specific place at each time, and no matter how small the division of time, you will never find the motion using this method. The authors point out that while we can often usefully imagine time as a sequence of points and motion as travel through those points, neither of those things is literally the case. As far as we know, the arrow travels continuously through space, and when we observe that the idea of individual points in time can never capture that motion, what we’ve located is a limit of the time is a series of points metaphor.

These kinds of complex puzzles are an interesting entailment of metaphorical thinking, and Aristotle was thinking about them over 2,000 years ago. In his discussion of place in Philosophy, he lists, “in how many senses one thing is said to be ‘in’ another.” Number eight is telling, and distinguishes this list from a simple recounting of a word used for eight entirely unrelated purposes: “In the strictest sense of all, as a thing is ‘in’ a vessel, and generally ‘in’ place.” In other words, Aristotle is describing and pointing out the literal, spatial meaning of the word “in” as the most fundamental key to understanding the word, along with a variety of metaphorical ways we can apply related concepts. Meanings one and two are still relatively physical: “1. As the finger is ‘in’ the hand and generally the part ‘in’ the whole,” and, “2. As the whole is ‘in’ the parts: for there is no whole over and above the parts.” This is a metaphorical dual, and Aristotle is pointing it out! Next, he goes on to describe two more similar meanings: “3. As man is ‘in’ animal and generally species ‘in’ genus,” and, “4. As the genus is ‘in’ the species.” Not only do these two meanings demonstrate a further level of abstraction, clearly showing how the original spatial metaphor can be applied to conceptual or categorical thinking, but the way that these two definitions mirror and repeat the first two also strongly suggests that the appearance of this dual can not be a coincidence, but a part of the intrinsic pattern of this metaphor that Aristotle observes repeating on different levels. Meanings five through seven are less metaphorically compelling, but include, “5. As health is ‘in’ the hot and the cold and generally the form ‘in’ the matter.”

A couple of pages of descriptions of ways that things can be in things later, he goes on to elegantly solve a Zeno paradox in terms surprisingly similar to Lakoff and Johnson’s: “Zeno’s problem — that if Place is something it must be in something — is not difficult to solve. There is nothing to prevent the first place from being ‘in’ something else — not indeed in that as ‘in’ place, but as health is ‘in’ the hot is ‘in’ the hot as a positive determination of it or as the hot is ‘in’ body as an affectation.” In other words, it’s a question of which way you interpret the spatial metaphor of one thing being “in” another. “So,” as Aristotle says, “we escape the infinite regress.”

**