Monday, September 27, 2004

Lakoff's View of Metaphors

In my last post on Lakoff, I mentioned that the aspect of his framing theory that I found the least useful for his political aims, and perhaps even detrimental to them, is his conceptual metaphor theory. I'm afraid that my reasons for finding this part of his framework unpalatable might not make sense to people whose only exposure to Lakoff's work has come through reading his political writings. While Lakoff has only recently become popular among liberal intellectuals, including those in the blogosophere, because of these writings, the theoretical paradigm from which his work on framing is derived is the product of more than twenty years of linguistic and pscholinguistic research by Lakoff, Mark Johnson, and others (most notably Mark Turner and Raymond Gibbs). Within this paradigm, the use of the word "metaphor" diverges in important ways from the everyday use of the word, and even from the use of the word in literary theories and contemporary theories of political discourse. While Lakoff summarizes conceptual metaphor theory in the first few chapters of Moral Politics, it's possible that many people who have not read his previous works on conceptual metaphors do not completely understand what Lakoff means when he says that our concepts are metaphorical. So, I thought it might be productive if I talked a little about conceptual metaphor theory, and what the word "metaphor" means to Lakoff and his colleagues.

The conceptual metaphor theory of cognition was first detailed by Lakoff and Mark Johnson in Metaphors We Live By. There they define metaphor in the following way:

The essence of metaphor is understanding and experiencing one kind of thing in terms of another. (p. 5)

This seems fairly normal, but the definition is built around a controversial thesis, namely that our conceptual systems are themselves metaphorical. Here is how they put it:

Metaphor is for most people a device of the poetic imagination and the rhetorical flourish--a matter of extraordinary rather than ordinary language. Moreover, metaphor is typically viewed as characteristic of language alone, a matter of words rather than thought or action. For this reason, most people think they can get along perfectly well without metaphor. We have found, on the contrary, that metaphor is pervasive in everyday life, not just in language but in thought and action. Our ordinary conceptual system, in terms of which we both think and act, is fundamentally metaphorical in nature. (p. 3)

They explain this view further with the following:

To give you some idea of what it could mean for a concept to be metaphorical and for such a concept to structure an everyday activity, let us start with the concept ARGUMENT and the conceptual metaphor ARGUMENT IS WAR. This metaphor is reflected in our everyday language by a wide variety of expressions:

ARGUMENT IS WAR
Your claims are INDEFENSIBLE.
He ATTACKED EVERY WEAK POINT in my argument.
His criticisms were RIGHT ON TARGET.
I DEMOLISHED his argument.
I've never WON an argument with him.
You disagree? Okay, SHOOT!
If you use that STRATEGY, he'll WIPE YOU OUT.
He SHOT DOWN all of my arguments.

It is important to see that we don't just talk about arguments in terms of war. We can actually win or lose arguments. We see the person we are arguing with as an opponent. We attack his position and we defend our own. We gain and lose ground. We plan and use strategies. If we find a position indefensible, we can abandon it and take a new line of attack. Many of the things we do in arguing are partially structured by the concept of war. Though there is no physical battle, there is a verbal battle, and the structure of an argument -- attack, defense, counterattack, etc. -- reflects this. It is in this sense that the ARGUMENT IS WAR metaphor is one that we live by in this culture; it structures the actions we perform in arguing. (p. 4)

The book, and much of their subsequent work, is full of examples like the ARGUMENT IS WAR metaphor, such as UNCERTAINTY IS UP, UNDERSTANDING IS SEEING, ANGER IS HEATED FLUID IN A CONTAINER, and LOVE IS A JOURNEY (see here for an extended list of conceptual metaphors). In fact, the primary (some would say only) evidence for their theory is the everyday use of statements that fit into these sorts of metaphorical expressions.

A second aspect of Lakoff's theory of conceptual metaphors is that the metaphors themselves are embodied. In other words, our concepts are built metaphorically from direct bodily experience. Embodied experiences that are repeated in our everyday experience (e.g., moving upwards) create what Lakoff calls "image schemas," which can then be used to structure less embodied experiences. Concrete, embodied experiences are therefore the least metaphorical, because they are built from direct experience, while more abstract concepts are structured metaphorically through mappings to more direct bodily experiences. UNCERTAINTY, a fairly abstract concept, is metaphorically structured by the bodily experience of "upwardness," and UNDERSTANDING, another abstract concept, from the bodily experience of seeing. Some concepts (e.g., LOVE) receive their structure from other concepts (e.g., JOURNEY) that are themselves structured metaphorically, from the image schemas of still more concrete (embodied) experiences.

There are several criticisms of this view, all of which together, in my mind, render it indefensible. The first, articulated by Greg Murphy1, notes that Lakoff and Johnson's almost exclusive reliance on linguistic evidence is both equivocal and circular (a common criticism of linguistic evidence in general). Is our speech metaphorical because our concepts are, our concepts metaphorical because our speech is, or is there some mediating factor? Lakoff and Johnson's analysis cannot distinguish between these possibilities. While there is some psycholinguistic evidence that our use and understanding of metaphors, idioms, and other types of figurative speech is connected to our conceptual structures, there is little or no empirical evidence that these concepts are, themselves, metaphorical. For Lakoff and Johnson's theory of conceptual metaphor to carry any weight, nonlinguistic evidence is needed. Twenty years into the conceptual metaphor project, it's hard to believe that this sort of evidence is forthcoming.

A second criticism, also from Murphy, argues that "structural similarity" between conceptual domains can account for the linguistic data without recourse to metaphor. Arguments are like wars in certain important ways, and this similarity, along with economical constraints on cognitive and linguistic structures, can explain the use of similar expressions in talking about the two domains. Nothing about the use of the expressions implies deep metaphorical relations between the two domains in our everyday representations of them. In essence, it appears that Lakoff and Johnson are doing what some refer to as "linguistic anthropology." Perhaps the terms used to describe arguments (or war) were once borrowed from the domain of war (or arguments), and thus used metaphorically. However, these once metaphorical terms have become lexacalized within the domain of arguments, and carry little or no connection to their former domain when used in the newer one. Once again, nothing in the linguistic evidence argues against this interpretation.

Murphy also faults conceptual metaphor theory for failing to explain the seemingly arbitrary choice of metaphors in everyday speech. Why do we use some experiences to metaphorically structure concepts, and not others which also bear structural similarity to the target concept? For instance, arguments could also be compared to journeys. Why aren't they? The reply to this criticism from the conceptual metaphor camp is usually that the direction of the metaphors, from more embodied to less embodied concepts, is important in determining what conceptual mappings are chosen. Thus seeing is used to structure our concept of understanding because seeing is a more embodied experience than that of understanding. This reply fails for many reasons, some of which I'll mention in a bit. For now, suffice it to say that the linguistic evidence alone fails to account for the particular metaphors that we find in everyday speech.

A final criticism from Murphy notes that many of the metaphors we use to describe the same concepts in everyday speech are inconsistent with each other. Is love a journey, or is it an opponent? Perhaps it is insanity, or a valuable commodity? Lakoff and Johnson observe that we use each of these metaphors in everyday speech, but Murphy shows that these metaphors often result in inconsistencies that would create concepts of love that contradict each other. This is perhaps Murphy's weakest criticism, as we now understand that concepts are not monolithic, static entities, of a single nature (see here for an argument against the "Natural Kind" view of concepts which views them as a single sort of fairly static representation). However, it does serve to further illustrate the arbitrariness of conceptual metaphor theory. It appears that the reason we have so many ways of talking about love is that, throughout the history of talking about love, people have used many different metaphors. Some of these metaphors have caught on, while others have not. The ones that have caught on have in turn become literal ways of talking about love, or dead metaphors, and the once metaphorical phrases have lost their connections to their original domains.

Two further criticisms not mentioned by Murphy are worth noting. The first involves the role of embodiment in conceptual metaphors, a role that is required for the theory to work (e.g., to explain the seemingly arbitrary choices of metaphors). In short, there is little evidence for this role. Obviously, we are embodied agents, and all of our conceptual representations are influenced by this embodiment, but there is no evidence that this influence is metaphorical, or that the relationships between concepts move in a single direction (from more embodied to less so). In fact, Lakoff's own example, from page 4 of Metaphors We Live By illustrates a conceptual relation that goes in the other direction! How many of us have direct, embodied experience of war? Very few of us, I would imagine. Still, the first and most prominent example used by Lakoff involves the concept ARGUMENT being structured by the concept WAR. I have a great deal of bodily experience with arguments, and none with war, yet I still talk about winning arguments, argument strategies, and defensive and offensive tactics in arguments. How can Lakoff argue that my concept of argument is structured through more direct, embodied experience of war, if I have none? Based on Lakoff's own analysis, I should never be able to learn this structuring of ARGUMENT, without the necessary bodily experience.

The last, and perhaps most important criticism of conceptual metaphor theory, particularly from the perspective of political theorists and practitioners who would want to use the theory to better frame political concepts, is that there it offers very little in the form of mechanisms. How are concepts mapped onto each other? How is structure from one concept carried to another? How are inferences from these structures made? How are these structures represented? Lakoff's theory offers no answers to these questions. Sure, he provides plenty of linguistic examples of the types of mappings (metaphorical, metonymical, polysemic, etc.), and even the types of inferences made, but no description of how any of this occurs. If we want to come up with new frames, and thus new mappings, we are going to need to have some idea of how these things occur. Without it, we are opening ourselves up to the kinds of problems that I mentioned in my last post on Lakoff.

1 Murphy, G. 1996. "On metaphorical representation", Cognition 60: 173-204.

Further Resources on Conceptual Metaphor and Related Theories


Lakoff on Conceptual Metaphor

Conceptual Metaphor

Women, Fire, and Dangerous Things by George Lakoff

Philosophy in the Flesh : The Embodied Mind and Its Challenge to Western Thought by George Lakoff and Mark Johnson

The Literary Mind: The Origins of Thought and Language
by Mark Turner

The Way We Think: Conceptual Blending and the Mind's Hidden Complexities by Gilles Fauconnier and Mark Turner

The Poetics of Mind : Figurative Thought, Language, and Understanding by Raymond Gibbs Jr.

Mappings in Thought and Language by Gilles Fauconnier

Unless you're really interested in doing research on Lakoff's conceptual metaphor theory, I would recommend skipping all but Mappings in Thought and Language and The Poetics of Mind. Philosophy in the Flesh is particularly bad, and philosophers who read it, especially those familiar with the last 100 years of continental philosophy, will probably find it downright offensive.

22 comments:

Bora Zivkovic said...

Thank you for this spot. I will link to it on my blog, for all those who do not understand what framing is all about.

Jim Kinney said...

I can appreciate your war argument as support for the fact that not all metaphors are rooted in embodied experience; however, the power of those metaphors to find emotional resonance and bind its users emotionally and psychically would certainly be greater if they had some visceral experience of things quoted.
Much of our experience is secondhand too. Traditons of storytelling and roleplay/theatre allow us to dabble in the what-ifs of scenarios whos visceral realities are merely inferred from vicarious experiences. Part of the function of language is to be a mutable platform of experiential exchange such that one can share in the experience of another without actually having "been there." It would seem that our ability to "suspend disbelief" in order to commune would invite the notion that we can, perhaps, make addendums to our own lexicon of experience using the narrative coding of another—a form of viral DNA insertion into host DNA that allows anothers lexical authority cohere to our own to such an extent that we can barely distinguish fact from fiction

Chris said...

Jim, I agree! In fact, that's one of the many reasons why figurative speech is so ubiquitous, and why so many different figurative expressions have become part of the language (in other words, have become literal). But Lakoff and Johnson don't really agree with you, or me, on that.

Anonymous said...

Very nice. Thank you very much. I’ve newly discovered Lakoff’s work on Conceptual Metaphors and it has blown my mind (Understanding As War metaphor? :-Þ).

In reference to politics I’d like to refer you to an old article of his, but one which still feels fresh & relevant to me in light of the 2nd (still ongoing) Gulf War:

Metaphor In Politics.

Anonymous said...

(If you could somehow incorporate my comments into one I’d really appreciate it!)

I meant to say that I’d newly discovered Lakoff’s works, so I enjoyed reading some counter–arguments. I shall definitely explore further.

Anonymous said...

One problem with Murphy's counterarguments is that there are actually between six to twelve different sources of support for the conceptual metaphor and embodiment hypotheses--and not just linguistic evidence. For example, in what is probably the most important book you left off your reading list, The Body in the Mind (1987), Johnson cites six bodies of evidence to support this claim, and in the Handbook of Cognitive Linguistics Rohrer cites 10 different bodies of evidence.


Similarly the objection that we have more experience of argument than war is overblown. Children play "war," not "argument." We play chess, the classic game of war, whose terminology overlaps with it, and we make sense of argument using both argument is war and argument is chess. Metaphoric structuring is always partial--that's one of the key points of the metaphor work. No concept that I know of, including abstractions like time, has no non-metaphoric structure.

Anonymous said...

Without being an expert myself, I would like to point out that some of your criticisms of Lakoff seem to be pretty outdated. In the epilogue to the 2003 edition of "Metaphors We Live By", Lakoff explicitly addresses some of these issues, and admits they were to some extent false.

Concerning ARGUMENT IS WAR, he suggests that it is an elaboration of an earlier ARGUMENT IS STRUGGLE metaphor. Besides, theories of Embodiment emphasize the importance of the individual in developing his own conceptual metaphors, so there is no reason to believe that ARGUMENT IS WAR is supposed to have a universal status. I.E. STRUGGLING TO INVENT AN EXCUSE IS ROWING A BOAT would be a metaphor that is highly culture specific.

Lakoff also acknowledges that the early theory underestimated the value of Primary Metaphors as being more fundamental than the later evolving Compunding or Conceptual Metaphors.

Finally, the point he addresses in this epilogue that seems most relevant to your versions of his theory is the "metaphor of metaphor". While his first conceptualisation of how metaphors work was in terms of 'mapping' (a topographical frame), it was later changed to 'projection' (deemphasizing the importance of transmission from source to target in favour of a third Bleded Space). Even more recently, the neural basis is invoked as a better model for understanding these processes.

With these later revisions in mind, I do not really see the relevance of the criticism you present here. And concerning the neural basis of language, Lakoff's theories seem highly compatible with the current investigation of temporal synchronisation of neuron ensembles as a way of binding percepts at a higher level (in contrast to lower, serially organized processing).

What are your alternatives?

Chris said...

First, let me just say that "ARGUMENT IS STRUGGLE" isn't a metaphor, it's a categorization! Arguments are struggles, at least much of the time, in a literal sense! To call it a metaphor is strange, and if some of the terminology from "struggle" is adopted for talk of arguments, that's probably because so many arguments are, literally, struggles.

It should be noted that while there is very little experimental or neuroscientific evidence for conceptual metaphor theory, there is absolutely no experimental or neuroscientific evidence for blending (a fact that Fauconnier and Turner fully recognize, but dismiss), so resorting to blends and the like (which I've discussed elsewhere on the blog) to defend an empirically falsified theory of concepts won't work.

Anonymous said...

Thanks for your reply!

Ok, I can follow your argument that describining "ARGUMENT IS STRUGGLE" as a metaphor may be difficult if metaphors are narrowly defined. However, to get over the merely terminological issues, let's change it into "VERBAL ARGUMENT" is "PHYSICAL STRUGGLE", which seems to be the underlying and rather obvious idea.

Following f.ex. Piaget's early suggestions with respect to cognitive development, it does not seem plausible that we come with innate categorisations, or the ability to categorize VERBAL ARGUMENT as an instance of MORE ABSTRACT STRUGGLE, but use the physical to structure the abstract. When you imply that arguments are literally struggles, where do you get the "neuroscientific or experimental" evidence for your assumptions from? I don't see how the lack of evidence from the former model can be a way for defening another model for which the same evidence lacks??

Substractive neuroimaging studies attempting to localize the alternative componential models are just as circular, by completely ignoring the developmental dimension.

Chris said...

Actually, I was speaking of the meaning of the words (which is where Lakoff gets his idea of metaphor in the first place), when I said that arguments can literally be struggles. Of course, there's no real evidence that we base our concept of "argument" on "physical struggle," even in the linguistic data, so I'm not really worried about that.

If you want neuroscientific data (why only neuroscientific?) on how categorization works, I'd be happy to give you a few hundred citations.

Anonymous said...

Ok, I'm not an American, so I am not going to argue with you about the meaning of the words. There might be some interference with the German and Scandinavian languages I am closer to, where ARGUMENT clearly belongs to the verbal domain.

Nevertheless, I still don't see how you avoid making assumptions when equating the "literal" meanings of the words to arrive at the conclusion that it is a matter of categorization rather than analogy.

Is there any evidence that words have literal meanings, apart from intuition and tradition? In this case, the frequently stated suggestion that we should be careful with attributing terms from one domain to another seems relavant in this context... F.ex. the genotype is often said to "be expressed", "to encode" certain traits ... or likewise, that words "have" meaning etc.

To me, Ramachandran's studies on synaethesia seem to provide strong suggestions that neural "cross wiring" is a fundamental aspect of cognition.

I guess a few hunders citations would be too much to ask, but a few would be nice in deeed! And the reference to neuroscientific or experimental data was merely quoting you. Feel free to include other disciplines : )

And, if categories are not embodied, developmental 'products'. where do you see a link to ontogenesis?

Anonymous said...

I don't like to be pushy, but I still don't see how the literalness of categories could be experimentally verified. In other words, I'm still waiting eagerly for a few of the hundred references you said you'd be happy to offer!

Chris said...

I'll give you two:
The Big Book of Concepts by Greg Murphy

Knowledge Representation by Arthur Markman

There will be hundreds of references is those.

You might also check out the work of any of the following:
Doug Medin
Brian Ross
Rob Nosofsky
Thomas Palmeri
Rob Goldstone
John Kruschke
Woo Kyoung Ahn
Steven Sloman
Bob Rehder
Todd Maddox
Greg Ashby
Evan Heit

Then check out the references in their papers on concepts and categorization.

Anonymous said...

Interesting that people are still stuck defending the classical notion that our categories are somehow fixed and external to ourselves, given to us in nature.

Why should argument necessarily be a struggle? As an academic and a lawyer, I've always found argument to be more of a dance than a struggle. Chris' emphasis on its "proper" categorization may reveal more about his tempermant than "the way it is."

The notion that a metaphor is a mere miscategorization is, in my view, an absurdist dance leading one right offstage.

Any abstract, high level concept is going to be governed by multiple embodied metaphors, and which one comprises our dominant "proper" categorization of a concept is going to be hotly contested within science and society. We choose the one that works best at the time on pragmatic, experiential and experimental grounds.

Jeff G said...

Very nice review Chris. I have to admit that as I read Metaphors We Live By I frequently found myself asking when the objection which you list would be responded to. While they did attempt to respond to some of them in their 2003 afterword, I couldn't agree with you more that their affirmative evidence seems to be a little lacking. It's nice to see somebody articulate many of the doubts and questions which had been swirling in their mind. Right now, I am working through their "Philosophy in the Flesh: The Embodied Mind" and am a little curious to see how much they will or will not rely upon their view of metaphors. I plan to do a couple of posts on the book, and your expertise will be more than a little appreciated.

Doug said...

I agree with some of the earlier commenters.

There is plenty of evidence for the embodied basis of cognition, language, etc. Our actions are the basis for cognition. I would suggest reading some work by for example Arthur Glenberg, Jean Decety, Bernard Hommel, Raymond Gibbs or browse any of the recent cognitive science society conference proceedings, they usually have at least a few embodied cog articles.

I don't know who Greg Murphy is. You didn't link to any publications of his. You tried to link to his blog it appears.

Here are some links to readings, references, and books that provide much more evidence in favor:

embodiment and cognitive science by raymond gibbs

the body in the mind by mark johnson

the cognitive neuroscience of action by jeannerod

where mathematics comes from by lakoff & nunez

(hci) where the action is by paul dourish

bernard hommel
http://psych.wisc.edu/faculty/bio/glenberg.html

art glenberg
http://home.planet.nl/~homme247/pubs.htm

Chris said...

Hi Doug, don't know if you'll actually read this, since this is such an old thread and you might not come back to it, but I thought I'd say a few things:

1.) Greg Murphy is a concept and categories researcher, and a widely known one; the link was supposed to be to the footnote, which is a reference to a paper of his.
2.)I'm quite familiar with the literature on embodied cognition, and while it is true that there is evidence for emobidied cog positions, it is limited evidence, and does not extend to conceptual metaphor theory.

Doug said...

Oh I see a Gregory Murphy who posted a response in the journal Cognition ten years ago (1996).

What can I say, a lot has changed in the past 10 years. There's not much point taking sides in a debate that happened 10 years ago. There is much much more evidence now in support of embodied cognition and the embodied basis for all forms of learning, metaphor, and other cognitive tasks.

I've seen similar pro/con journal articles and issues about Gibson's ecological psychology (Perception), situated cognition (cog sci), and others. You can't take the simplistic view that one side "won" or another side was rendered "indefensible", especially when it was written over a decade ago.

Doug said...

And I forget to mention Jean Decety's publications (most of which came after the 1996 debate):
http://home.uchicago.edu/~decety/publications.html

Chris said...

Doug, just so you know, I'm well aware that much has gone on since the 1996-97 debate between Murphy and Gibbs (the first article in the debate, by the way, was Murphy's), but most of Murphy's arguments are actually theoretical, rather than empirical, and they still stand. I recommend reading it.

I should note that I am actually a participant in this debate, both theoretically and empirically, so I am pretty confident that I know where it stands.

Anonymous said...

I haven't read all the comments on your entry, so I'm nor sure if somebody else has pointed that out.
In respect to your argument "Based on Lakoff's own analysis, I should never be able to learn this structuring of ARGUMENT, without the necessary bodily experience."
Lakoff (1999) says that metaphorical mappings either result from pre-conceptual embodied experiences or build on these experiences and general knowledge in order to construct more complex conceptual structures. So, the embodied experience is only true for primary metaphors like MORE IS UP which, together with general knowledge and cultural models, motivate more complex metaphors (cf. Lakoff and Johnson 1999: 57, 60). So, in my opinion, to understand the concept WAR, we do not have to experience it directly. I also think that an argument is more abstract than war is. In war, e.g., you can actually see bombs being dropped etc. In an argument you "fight" with words, which I think is a little more abstract. You could also see a discussion of a similar metaphor LOVE IS A FINACIAL TRANSACTION in "Figurative language" from Dmitriĭ Olegovich Dobrovolʹskiĭ, Elisabeth Piirainen (p. 139, 140). And maybe an article by Gallese and Lakoff (2005)in "Cognitive Neuropsychology" would also be interesting in which they talk about the neural representation of concepts.

Anonymous said...

Anothre set of ralph lauren polo in three colors. Ralph

Lauren Polo Shirts
come with a graduation of colors on the chest. The new shiki Cheap polo

shirts logo is on the chest. Comes in all layres and can be worned tucked in or out Wholesale Polo Shirts -50% OFF.
Also comes with sculpted collar and arm cuffs.Flat knit collar, Contrast placket and half

moon, Mothre-of-pearl buttons and tennis tail, Soft double knit piqué,100% cotton ralph lauren

polo shirts. If you lend your embroidrere a sample cheapralph lauren polos that will help you get

the closest match to the original embroidreed design. The artwork you provide will sreve as a

template for your embroidrey set-up but discount ralph lauren polos can't be used to identify some

stitch types, cheap polo shirts Cheap Polo Shirtsgive

precise sizing of embroidreed elements and show thread polo clothing colors.

Anothre set of ralph lauren polo in three colors. Ralph

Lauren Polo Shirts
come with a graduation of colors on the chest. The new shiki Cheap polo

shirts logo is on the chest. Comes in all layres and can be worned tucked in or out Wholesale Polo Shirts -50% OFF.
Also comes with sculpted collar and arm cuffs.Flat knit collar, Contrast placket and half

moon, Mothre-of-pearl buttons and tennis tail, Soft double knit piqué,100% cotton ralph lauren

polo shirts. If you lend your embroidrere a sample cheapralph lauren polos that will help you get

the closest match to the original embroidreed design. The artwork you provide will sreve as a

template for your embroidrey set-up but discount ralph lauren polos can't be used to identify some

stitch types, cheap polo shirts Cheap Polo Shirtsgive

precise sizing of embroidreed elements and show thread polo clothing colors.
Wholesale Polo Shirts
Ralph Lauren Polo Shirts
Discount Ralph Lauren Polos