Intro to Metaphor Theory

If your experience was like mine, when you learned about metaphor in school, the lesson was mostly about remembering the difference between metaphor and simile. (Similes make analogies explicit by using the word “like” or “as;” metaphors don’t.) This turns out to be among the least important dimensions of how metaphor works. Instead, theorists try to work out (and disagree over) how metaphors structure how we think about and work with their targets. The four approaches on this page summarize major differences among schools of thought about how metaphors work in practice, though individual researchers have described innumerable variations on each theme.

These four approaches aren’t simply a historical progression; you can find scholars espousing each of them today, and the second two were around long before they became popular. Nevertheless, to some extent, the first two more “conservative” or restrictive stances have been supplanted by the latter two more “progressive” or expansive ones.

Metaphors as poetic devices

The narrowest way to understand metaphor is as an expressive or poetic way of speaking, used for dramatic or emotional effect. Someone employing this lens sees metaphors primarily in creative writing, and in expressive or “flowery” speech. They’re seen as important to how language affects its audience, but not essential to the action; you can remove the metaphor and still understand what’s going on, even if the results aren’t as emotionally affective or entertaining. Most scholars now accept that metaphors are far more widespread than this position suggests.

Metaphors as part of everyday language, but not ideal

You’ll find this view in the entry on metaphor in the Stanford Encyclopedia of Philosophy (a great resource in general, though its articles often take a narrower or “less objective” view than you’d expect for an encyclopedia), where the authors continually talk about “resorting to metaphor.” They imply a distinction between ideal circumstances, when you can just say what you mean, and less-than-ideal ones, when you have to “resort to metaphor” to express yourself or make your point. At the top of the entry, they define metaphor as “a poetically or rhetorically ambitious use of words,” again suggesting that something out of the ordinary is going on when metaphors show up.

According to this way of thinking, metaphors are things to avoid because they’re (to keep quoting the article), “rich and unpredictable”—they can do lots of things, some of which may be hard to predict. From a positivist perspective, that’s a problem because the goal of language is to describe things as neutrally and precisely as possible—a perspective that makes sense if you believe that non-metaphorical or “ordinary” language can precisely indicate one thing, and that it is possible to know something independent of language. Many rhetoricians would disagree, arguing that all language is more or less ambiguous and capable of multiple meanings depending on the context in which they’re read. Some would also point out that we can only know things through language, so that the idea of using language to precisely point to one thing that exists outside of language doesn’t make sense, though scholars hold many different views on that point.

Metaphors are essential and everywhere, but easily distinguishable from “ordinary” language

This view was widely popularized by a book called Metaphors we Live By, written by George Lakoff and Mark Johnson in 1980. According to Lakoff and Johnson, metaphors are everywhere in everyday language. Far more than simply expressing emotion, metaphors structure language-users’ thoughts about ordinary, everyday things. People afflicted with cancer are “fighting a war” when they undergo treatment, for example, while the government fights a war on cancer by funding cancer research. North is up, and up—like other “positive” and “on” things—tends to be good; feeling up is better than feeling down. “Dead” metaphors, or metaphors that aren’t generally recognized as metaphorical by the people using them, such as going “up” to Alaska, still structure everyday thinking in this way.

This approach suggests that metaphors in everyday language can encourage ways of thinking that aren’t aligned with the goals and purposes of the person using them. For example, research has found that war metaphors make some people living with cancer feel discouraged rather than emboldened because their only option is to win or lose, because there’s nothing they can do to stop war from happening, or because they don’t think of themselves as fighters. War metaphors are also gendered, in that more men fight in wars than women, and because in most cultures, women are encouraged to be less violent and aggressive than men.

Lakoff and Johnson popularized a way of talking about metaphor structure that makes it easy to talk about these patterns, and that many researchers employ today:

Source domain —> target domain

The source domain is the source of the imagery in the metaphor, e.g., war.

The target domain is the object to which that imagery is applied, e.g., cancer.

A METAPHOR, in Lakoff and Johnson’s terminology, is the big overarching between the source and target domains, e.g. CANCER IS WAR. Conventionally, metaphors are written in all caps to make them easy to distinguish from metaphorical expressions, or ways of talking that fit under the parent or umbrella metaphor. When you call someone undergoing chemotherapy “a fighter,” or talk about “invading” cancer cells, you’re using metaphorical expressions under the parent CANCER IS WAR metaphor. This distinction is useful because many different metaphorical expressions may all be part of (and reinforce the way of thinking encouraged by) the same metaphor.

Metaphors essential and everywhere, with fuzzy edges

This approach—the one we take in this resource—observes that at some level, all language is metaphorical, so that “literal” and “metaphorical” aren’t binaries. Many expressions can’t simply be labeled as one or the other. Instead, language is always about comparing one thing to another, constructing categories so that the same word can be used for things that are, strictly speaking, different. Investigating metaphor is about understanding what those comparisons do.

Lakoff and Johnson’s approach assumes that distinguishing between literal language and metaphorical language is trivial. This is an unhelpful assumption! Language changes over time. An expression that may be understood metaphorically in one time and place may be understood literally in another. The same expression or phrase may also be understood differently across contexts or discourse communities (communities who share a discourse, a way of communicating). Molecular biologists may understand themselves to be “literally” editing the genetic code, while “editing the genetic code” or “editing DNA” may look like a metaphor in the popular news.

And who has the authority to decide whether something is or isn’t a metaphor, anyway? In the vast majority of cases (French as spoken in France is an exception, or at least the Académie Française responsible for policing the official version of the language thinks so), no one has the authority to define what a word “actually” means for every use and user of that word. That goes for scientific language, too. Some researchers may assert that certain definitions are correct or incorrect, but doing so only enforces that definition if those researchers are part of a standards-setting body, and then their definition is only authoritative for work done under that standard.

All language can be understood as metaphorical in the sense that every time you use a word for a thing, you’re likening it to other things for which you’ve previously used the same word. You’re making an argument for a particular way of seeing it, a judgment about which kinds of similarities and differences are most important. In some instances, the choice may not seem like much of a choice because everyone (everyone in the relevant discourse community!) shares the same norms around those judgments. Yet even if an individual language-user can only realistically or reasonably make one choice, other choices always exist in the broader sense of how social norms could be otherwise.

A chair is a chair because its similarity to other things we’ve previously learned to call chairs is strong and those chair-ish characteristics are obviously relevant to how it’s used. A car might be a car or a vehicle or a sedan or a Volkswagen or a junker or a ride or a baby; each emphasizes different similarities to a different group of other “like” things, in ways that may be useful to in particular contexts.

DNA can be discussed as a polymer, a genetic code, a blueprint, instructions, genes, and so on. We might say that DNA “literally is” a polymer, but each choice highlights what kind of thing DNA is in a different way, through comparisons to different sets of “like” things.

Following this logic, metaphors are ubiquitous and unavoidable in specialized scientific language, because communicating science involves:

• Metaphors prevalent in non-specialist or “ordinary” language that also happen to show up in science, e.g. referring to performing a set of activities as “conducting” an experiment
• “Dead” metaphors incorporated into a field’s standard jargon, for example, referring to membrane-enclosed units of living things as “cells” (originally an obvious metaphor comparing the way those units appear under a light microscope to a monk’s or nun’s cell) or “running” a gel (flipping a switch that allows electricity to flow or “run” through an apparatus for performing gel-electrophoresis)
• Explicitly metaphorical language that’s essential for describing new (or relatively less well-understood) phenomena in terms of more familiar things, just as explicit metaphors are used outside science, for example, describing CRISPR as “gene editing” (relying on the parent metaphor DNA IS TEXT to explain changing a base pair in terms of changing a letter in a word)
• Constantly comparing things to other things across all, as we’ve explained above.

Further reading

• For more on how language structures thought, see the work of Lera Boroditsky