Analogy and Metaphor

Lakoff and Gentner

Posted on September 24, 2009 by Peter Turney

In response to my previous post, Criticisms of Lakoff’s Theory of Metaphor, Cosma Shalizi suggested that I should look at some criticisms from Chris and Murphy:

Here are my thoughts on these criticisms.

This is what I make of Murphy’s paper: What Murphy calls Structural Similarity (Sections 3.3 and 6) is the right approach to metaphor. Murphy’s idea of Structural Similarity is that (something like) Gentner’s Structure Mapping Theory (SMT) [1] is the underlying mechanism for Lakoff’s Conceptual Metaphor. I support this view [2], as does Gentner [3,4,5]. The other views Murphy talks about, strong and weak metaphoric representation (Sections 3.1 and 3.2), are strawmen.

However, it seems to me that Murphy does not fully understand Gentner’s work. In Section 3.3, he says, “The structural similarity view is not a theory of novel verbal metaphor.” This is wrong. The structural similarity view is exactly a theory of novel verbal metaphor [3]. Murphy doesn’t seem to understand that the point of Gentner’s SMT is to transfer knowledge from one domain to another. Murphy seems to believe that the two domains (source and target) must be completely structured before SMT can be applied, and that SMT does not add any new structure. If this were so, then the mapping process in SMT would be no more than an entertaining exercise, rather than a method to generate new knowledge.

Summary: the bulk of Murphy’s paper is spent on presenting and refuting strawmen. His proposed solution is the right way to go, but he doesn’t fully understand his own proposal.

[1] Structure-mapping: A theoretical framework for analogy, Gentner (1983)

[2] The latent relation mapping engine, Turney (2008)

[3] The career of metaphor, Bowdle and Gentner (2005)

[4] Metaphor is like analogy, Gentner, Bowdle, Wolff, and Boronat (2001)

[5] Metaphor as structure-mapping, Gentner and Bowdle (2008)

In Explaining the War of the Metaphors, Chris writes:

“I and others think that CMT and blending would benefit greatly from research on analogy. If these theories had a testable model of how mappings between conceptual domains are formed, interpreted, and used — a model that structure mapping theory can provide — it would be possible to make some concrete, experimentally testable predictions about things like inference, object recognition, etc.”

I completely agree that CMT (conceptual metaphor theory) would benefit greatly from research on analogy; in particular, the research of Gentner and her colleagues.

In Idioms, Metaphors, and Lakoff, Oh My!, Chris writes:

“Sometimes I forget that not everyone who happens upon this blog today has been reading it from day one (I mean come on, why haven’t you?). It surprises me, then, when people tell me they’ve seen no evidence that George Lakoff and Mark Johnson’s conceptual metaphor theory is, well, wrong.”

I disagree that Lakoff and Johnson’s theory is (wholly) wrong. Some details may be wrong, but I think that many of the core ideas are correct and furthermore consistent with Gentner’s theories and experiments.

It seems to me that Chris is arguing two incompatible claims: (1) Gentner’s theory is the proper foundation for Lakoff’s theories. (2) Lakoff is wrong. (If not incompatible, these claims are at least dissonant.)

In Metaphor as structure-mapping, Gentner and Bowdle write:

“We have suggested that metaphor is like analogy — that the basic processes of analogy are at work in metaphor. Specifically, we suggest that structural alignment, inference projection, progressive abstraction, and re-representation are employed in the processing of metaphor and simile. This view can help resolve some tensions in the field: for example, on this view, metaphor both reflects parallels (Murphy, 1996) and creates new similarities (Lakoff, 1990) between the domain compared, via structural alignment and candidate inferences, respectively.”

I agree with Gentner that her view “can help resolve some tensions in the field”. It puzzles me that Chris simultaneously agrees with Gentner and yet seems to actively work to increase tensions in the field. Instead of attacking Lakoff, why doesn’t Chris work to integrate Lakoff’s and Gentner’s theories, as I attempt to do?