Louder than words: The new science of how the mind makes meaning. By Benjamin K. Bergen. New York: Basic Books, 2012. Pp. 312. ISBN 9780465028290. $27.99

Imagine that you are a participant in the following psycholinguistic experiment. You are seated in front of a computer terminal and shown the sentence The carpenter hammered the nail into the wall. After reading the sentence, you are shown a picture of an object, such as a nail or elephant, and asked to quickly judge whether that object was mentioned in the sentence. Of course, you would quickly say ‘yes’ to the picture of a nail and ‘no’ to the elephant. The primary interest, however, is in your speeded response to the nail picture, depending on whether it was shown in a horizontal or vertical orientation. Research indicates that people, on average, are faster to make their ‘yes’ decisions when the picture was in the same spatial orientation as implied by the sentence just read (Zwaan et al. 2002). Thus, people are faster to say ‘yes’ when the picture showed the nail in the horizontal orientation than when it was shown upright, or in the vertical position. However, when they first read the sentence The carpenter hammered the nail into the floor, people are faster, on average, to say ‘yes’ to the nail picture that presented it in a vertical position rather than horizontal.

One interpretation of these findings is that people automatically construct a mental image of an object in its appropriate spatial orientation based on what the sentence implies. Even if the nail’s position is not explicitly noted in the sentence, our immediate understanding of the sentence’s meaning enables us to create an image of the situation in which the nail was hammered in a horizontal or vertical position. How people construe imaginative understandings of language is the subject of Ben Bergen’s book. This lively, entertaining book offers a broad, but detailed, overview of the idea that people interpret language using embodied simulation processes. This hypothesis asserts that people ordinarily construct imaginative reenactments of what some language event must be like to participate in given their own bodily capacities and experiences. People do not first comprehend a sentence’s purely linguistic meaning and only then derive richer imaginative understandings. Instead, people’s embodied simulation abilities enable them to immediately infer detailed, imagistic understandings of what speakers imply by what they say.

The embodied simulation hypothesis has been widely debated within the cognitive science community in recent years as it possibly applies to various cognitive and linguistic phenomena, including conceptual representations, memory, problem-solving, learning, and consciousness. B’s book progressively outlines the arguments and empirical evidence for embodied simulations in terms of how people make sense of linguistic meanings. Ch. 1 presents the main idea of embodied simulation processes by showing how people can imagine unrealistic scenarios, such as ‘flying pigs’, by combining mental representations of different percepts that they have experienced (e.g. experiences of pigs and flying). B contrasts this view of language understanding with a traditional account in which words are comprehended by looking up their definitions in a mental lexicon, a repository that is assumed to be quite distinct from embodied experiences and actions.

Ch. 2 describes how many facets of mental imagery are closely tied to the ways our brains move our bodies and perceive the world. One key finding is that mental imagery sometimes interferes with visual perception. For example, when you form a mental image of a banana while looking at a blank wall, this hinders your ability to perceive a faint image of a banana projected onto the wall. Various cognitive neuroscience studies demonstrate that brain areas responsible for visual perception are also engaged when people are only imagining some object or scene or remembering some past action (e.g. making a fist). These different experimental results emphasize both the tight coupling of perception and action, and the possibility that people use their perceptual and motor systems for simulation purposes.

Ch. 3 begins the discussion of the large body of experimental research within psycholinguistics on embodied simulations, such as the earlier-mentioned studies showing how people automatically infer the spatial orientation of a nail when reading The carpenter hammered the nail into the wall. B does an excellent job leading readers through these complex experiments and highlighting the need to always consider alternative hypotheses within any set of experimental studies. [End Page 531]

Ch. 4 goes on to describe the well-known studies on mirror neurons that show that looking at some action activates brain areas relevant to performing those same actions. Embodied simulations, which recruit the mirror-neuron system, enable people to project themselves into the minds and actions of others, including the objects and events referred to as people talk. For example, studies on the ‘action-compatibility effect’ indicate that people are faster to make comprehension responses for sentences like John opened the drawer when they have to move their hands toward their bodies to push a comprehension button than when they have to move their hands away from their bodies (Glenberg & Kaschak 2002). The reverse pattern of results is observed when people hear sentences implying movement away from the body, such as John closed the drawer. Once again, people interpret sentences by imagining themselves engaging in the very actions specified in the language.

Ch. 5 explores the role that grammar plays in embodied simulation processes. For instance, one set of studies employs the method used to establish the ‘action-compatibility effect’ to show that grammar shapes simulation processes (Bergen & Wheeler 2010). In these studies, people read either progressive sentences, such as John is closing the drawer, or perfect sentences, such as John has closed the drawer. Participants are faster to make their comprehension responses for progressive sentences than perfect ones when they move their hands away from their bodies. Thus, interpreting perfect sentences evokes fewer embodied simulations of movement compared to the progressive sentences, which require more detailed simulations to properly understand. B goes on to offer several intriguing possibilities for the different ways that grammar may modulate on-line simulation processes.

Ch. 6 is devoted to showing how embodied simulations are constructed incrementally during speeded sentence comprehension. Consider the sentence in 1.

1. 1. Before / the / big race / the driver / took out / his key / and / started / the / car.

The implied direction of the key turn in this case is clockwise. In one experiment, participants read through each sentence by rotating a knob after each chunk of words, indicated by the slashes in the above example (Zwaan & Taylor 2006). Participants turned the knob in either a clockwise or counterclockwise direction. The result of interest here was that people are faster to comprehend the verb started when they make their knob turns in a clockwise direction than when making counterclockwise rotations. People essentially understand the key verb started by constructing an embodied simulation of the implied movement the car driver had to do in order to turn the key and start the engine. This illustrates that people do not have to wait to the end of the sentence to initiate their simulation processes. Embodied simulation processes are not optional, after-the-fact operations that emerge only after a sentence has been read and understood.

Not surprisingly, people with different experiences and expertise may vary in how and when they construct embodied simulations. Ch. 7 describes several relevant experimental studies in support of this idea. Ch. 8 outlines how people’s experiences with different languages also influence how they interpret language. For example, languages like Arabic or Hebrew are written right to left; others, such as English and Italian, are written left to right; and still others, such as traditional Chinese, are written top to bottom. In a study by Maass and Russo (2003), when Italian or Arabic speakers heard sentences like The girl pushes the boy, and then judged whether a picture properly captured the event, the participants responded differently. Arabic speakers, for instance, took less time to judge a picture as correctly depicting the event when the girl was on the right side of the frame, while Italians took less time to make the same judgment when the girl was shown on the left side. Thus, people’s experiences with the spatial direction of written language affects the embodied simulations created when they interpret simple action sentences.

Ch. 9 considers the embodied simulations that arise when people interpret metaphorical language. Critics of embodied simulation wonder whether such simulations are even possible when thinking about abstract ideas and topics. Studies show, however, that having people first make a hand movement, such as reaching out to grasp something, facilitates the speed with which participants subsequently comprehend a metaphorical phrase such as grasp the concept (Wilson & Gibbs 2007). Even if people are unable to physically grasp a concept, engaging in relevant body actions primes the construction of an embodied simulation to infer the metaphorical meaning of [End Page 532] the phrase. Neuroscience work has also shown activation in the motor system of participants’ brains when they read both literal (e.g. grasped the stick) or metaphorical (e.g. grasped the idea) statements (Desai et al. 2011), which offers additional evidence that embodied simulations may underlie our understanding of metaphorical meanings.

Ch. 10 discusses some of the functional advantages that embodied simulation processes have in language use, and strongly argues that these processes are not, once again, optional ‘downstream’ activities, but are critical aspects of linguistic interpretation from the early stages of processing. B suggests, though, that it is not yet entirely clear if embodied simulations are necessary or sufficient for understanding linguistic meaning even if the empirical research strongly shows that they play a functional role in how people make sense of many aspects of language.

Ch. 11 raises several speculative questions about how embodied simulations operate and what their possible existence implies about the very nature of human communication. Finally, the epilogue describes the ‘crosstalk hypothesis’, which attempts to explain why people sometimes have a difficult time driving and talking on a phone at the same time (i.e. the simulation processes used for interpreting language may interfere with the perceptual/motoric actions you need to drive). In this way, embodied simulation is not just observed in the confines of psycholinguistic laboratories, but also is relevant to human action in various real-world contexts.

There are only a few criticisms of Louder than words that are worthy of mention. First, B could have articulated a more expansive story about traditional theories of linguistic comprehension, most of which embrace disembodied approaches to language, to more sharply emphasize the revolutionary nature of the experimental findings on embodied simulations. Second, more could be said about the roles that embodied simulations may play in other aspects of human experience, ranging from perception (e.g. we perceive objects by imagining the possible bodily actions we could engage in with those objects) to consciousness (e.g. we consider possible future actions through embodied simulations of what those actions may be like to engage in and what they may produce). Finally, I personally believe that B could have been more forceful in defending the embodied simulation hypothesis when outlining several possible limitations of it in Ch. 10.

But these slight criticisms do not subtract from the pleasures of reading this book and coming to appreciate its many lessons about the importance of embodied simulations in how people create meaningful interpretations of language. B has provided a major theoretical statement, backed by a large assortment of experimental research, on the embodied foundations of linguistic meaning and on how we automatically understand language in real time given our abilities to imagine ourselves engaging in the very actions we are hearing or reading.