Review of Benjamin K. Bergen's "Louder than Words".

Benjamin K. Bergen’s book “Louder than Words” has gained a lot of attention as a work of popular linguistic theory, combined with neuroscience. I enjoyed reading the book a lot, and I felt that it was a very insightful work for the “linguistic layperson”, as well as perhaps for the neuroscience “layperson”, and it could easily help said people to get an introduction to some very fascinating research dealing with the use of language in both a quantifiable and objective format. Or, put in other words, this newly-published book deals with what science can objectively say about how human beings understand language. Most people, especially those people who have little to no knowledge of studying either their own language or a foreign language, probably are not very curious about how people use their brains to understand language; indeed, language is such a ubiquitous part of our everyday experience as people, that most of us tend not to even give it a second thought. We talk, we write, we read, we listen to what other people have to say to us, and we focus on the subjects of the incoming words that we process. However, have you ever stopped for a minute to consider how you process all this incoming information? Furthermore, have you ever even classified the words that you hear and see, or form in your own head, as “information?” According to the way that a neuroscientist thinks, living beings (and this includes every living creature) react to stimuli (or perhaps one could just call it external stimulation). Creatures with complex nervous systems react to the things that they perceive by sight, sound, smell, and sensation. Though as humans, we do not like to deconstruct our senses and call what we perceive around us “stimuli”, for the sake of science, we are forced to do so if we wish to remain objective. Thus, we have learned to not merely react to things that we see around us, but human intelligence, with the help of language, has enabled us to think of the things that we see and hear around us as “stimuli”, and also to have a constant internal dialogue within ourselves to which we also react. These concepts of stimulation and reaction essentially form the core research of neuroscience. This research is then used in order to map the brain, and determine which areas of the brain are used for specific functions. This process of mapping the brain is undertaken by scientists with the help of brain scans, and a machine that measures neural oscillations—otherwise known as brain waves. Up to this point, the thing that had hitherto not been attempted, however, was to determine whether or not talking and hearing about specific mechanical actions would actually fire certain parts of the brain that are responsible for the actual execution of various mechanical functions of the body. For example, Bergen and his researchers wanted to know whether or not the thought of pounding a nail into a board would actually fire the specific neurons in the brain that activate the muscles in the wrist used for pounding. What they discovered was that, yes, merely thinking thoughts about actions undertaken by the limbic system was enough to activate the same brain regions which control the actual physical actions themselves.

            The thing that Bergen’s new book relentlessly goads his readers into considering is that, before completing an action such as pounding a nail into a piece of wood, or imagining the color of a polar bear’s nose, as people, we are required to use language and grammar in order to accomplish the task of imagining something. However, it is not spoken language, but rather, something that he labels as “mentalese”, or the sentences in one’s head that one must think, which is transferred to an action which is to be carried out either immediately, or at an indefinite time in the future by the specific human who is to perform the specific action. It is true that other mammals such as dogs and chimpanzees can successfully memorize hundreds, and in some cases, thousands, of new words, but the way that humans use language, we can imagine entire scenarios in our minds, and plan out what we will do before we do them, as well as engage in complex discussions with other people about innumerable topics. A parrot can, for instance, learn to recognize a toy and say what the object is to its owner; the same parrot can probably also figure out how to open a cage on its own, as well communicate some bit of parrot information to another parrot. However, a parrot cannot use language to figure out how to get out of its cage, and then explain to its owner what it did in order to escape. The fact that human beings can use language to plan, and also communicate complex ideas, is one of the key traits that separate us cognitively from other animals—even other animals that are exceedingly intelligent and perceptive. Indeed, our collective ability to listen to instructions, follow them, and also to create our own plans and execute them within our own minds, is one of the things that seems to indicate that we actually do have free will to do what we want, as opposed to being confronted with a stimulus, such as a predator, and merely reacting to it by running away, or fighting back against it. This is, of course, not to suggest that other animals don’t have the capacity to reason; however, it should be an indication that humans are the only animals that seem to have a complex inner-life of emotions and plans reverberating within the confines of their skulls. The reason that we do have such complex thoughts, and that on a given day we can make a “to-do list”, is precisely that we can use language in the way that we do (animal language is a very fascinating topic, but I am a linguist, and not an animal behaviorist, so I will not be able to offer any insight whatsoever on said topic). The specifically human pattern in which we use language—which more or less is to gather information about one or multiple topics and either keep track of them, or act on them, is accomplished with the help of 2 things, without which we could not create any thoughts: grammar, and syntax. Within the field of linguistics, grammar is defined as something that exists in all human language, which ends up being the sum total of everything within a given language—from how words are formed, how actions are spoken of in the present tense, all the way to how objects are taken from being distinguished between a singular and a plural form. Syntax, on the other hand, can generally be defined as the proper word order of a given sentence within a language.

            Back to the subject of Bergen’s book, the research that he and his fellow University of California-San Diego researchers performed consisted mainly of testing how both grammar and syntax have an effect on test subjects’ response to reading and hearing sentences about performing actions, as well as imagining various scenarios and objects in their minds. In a nutshell, Bergen’s hypothesis about language and its relation to the motor system is that humans, essentially, have to perform what he defines in his book as an embodied simulation about everything that they hear or read via a visual process in their minds, and that individual words are merely verbal cues that activate simulations related to limbic and other motor systems in the brain. He further hypothesizes, essentially, that the cerebral cortex is co-wired with particular words to activate particular actions, in different brain regions. This would mean, for instance, that there is a set “what” and “where” pathway for neurons on which to run, and that by a more or less “Pavlovian” association of verbal or written cues, which we associate with our direct interactions in and our experiences of the physical world which we have already experienced. What is linguistically new about his hypothesis is that he and his researchers believe that the “mentalese” that we all speak to ourselves within our own minds is very much dependent on our own motor systems, as opposed to the traditional view that grammar and syntax are more or less independent entities that can be used freely by the human mind to think of any thought that one’s heart desires on a whim. Throughout the book, Bergen makes it fairly clear that he is placing a lot of hope into his hypothesis, as he believes that a “stimulus/reaction” scenario is much more in-line with what is already known about evolution as opposed to a theory of a sort of grammar that is independent of visual and other sense-associations, and that by looking at our current brain-language machinery, we can easily make some leaps and fill in some of the gaps in the theories as to why humans’ highly advanced intelligence with regard to articulated abstract thoughts and sense of freedom of action does not seem to make sense within the framework of what I will call for the sake of this essay “the stimulus-reaction paradigm”, the paradigm which is more or less the basis of modern neuroscience and neural evolutionary theory. In short, he is postulating that our ongoing internal dialogue within ourselves must be, by the framework imposed by what we know about evolution, more or less an illusion, and that even our capacity to think abstractly is based on things such as spatial relationships and deductions based on complex yet entirely subjective associations with individual words that the observer makes.

            Another important and extremely challenging point that Bergen makes with his work is that individual words, and thus not syntax and grammar, must directly trigger the same physical response as does actually physically completing the same action that said words represent—when a test subject is having his/her brain either scanned or his/her brain waves measured. And not only does Bergen predict that syntax and grammar are themselves not responsible for firing specific neurons related to specific motor actions within the cerebral cortex, but furthermore, that other parts of speech within a sentence (besides nouns and verbs), such as tense markers, conjunctions, adverbs and the pronoun markers that denote the idea of “person” (the difference between I, her, and y’all), are all based on subjective impressions of spatial relationships that the respective interpreter of a sentence is making between physical objects, and also abstract concepts. Or, in other words, individual words must trigger associations that a person makes about his/her physical body with both physical objects and abstract concepts, and that the language of thought within our heads must therefore be an illusion—something that is physical at its core, and not detached from the stimulus/response paradigm at all. For instance, an experiment within Bergen’s research indicated that test subjects having their brains monitored for activity pressed buttons in such a way that indicated that they perceived events described as having occurred in the past tense in the same way that they perceived an object that had been described as being far away from them, and that they had also pressed a button in such a way that indicated that they had perceived events described as having occurred in the present progressive tense in the same way that they had perceived an object that had been described as being physically close to them. Bergen writes about his own research:

                        “The experimenters found that participants responded to completed-state

                        pictures faster than ongoing-state pictures following perfect [past tense]

                        sentences. This suggests that the participants were not representing the

                        internal structure of events described with the perfect, so much as their

                        resulting endstates.”

                                                (Louder than Words, pg. 116)

What this seems to imply, therefore, is that the grammatical structure representing tense is not so much evidence of human beings’ using language to grasp abstract concepts, as it is to imply that a structure such as tense is a word that triggers a sort of “Pavlovian” association that regulates spatial relationships in the mind of the respective speaker, and that the grammatical structure itself is more or less an illusion.

            The results of Bergen’s research are, however, inconclusive, as much as he and his researchers appeared to be in favor of its being true throughout the book. However, Bergen readily admits that there is not enough evidence to come to any one conclusion or another about it. The conclusion from his research that can be conclusively made, however, is that the new “neurolinguistic” concept of embodied simulation is indeed a real phenomenon, and that it does play a significant role in a person’s mental process of understanding all kinds of situations, learning how to do new tasks, and performing a specific physical action. And it is indeed quite fascinating to contemplate the fact that human beings interact with the world, and form associations about it, based very much so on a sort of “personalized experience”—which is to say that we as human beings are in a certain way “Pavlov-ian”. What I believe to be true about such a “personalized experience” view of human language is that we do partly use our physical bodies while we are speaking language, and that our mental associations are probably a large piece of the puzzle when it comes to our ability to acquire language in the first place. But to jump to the conclusion that the way that we understand language must be entirely “Pavlov-ian” is a bit of a leap too far, and in my opinion, it does not do human language its due justice. The logical conclusion that could be reached by the embodied simulation hypothesis, when taken 100% literally, is that we do not, as people, really experience grammar at all; rather, we associate certain words with outside actions in a completely visceral way. Therefore, we must come to the logical conclusion that language is nothing more than another (albeit a currently inexplicably complex one) stimulus/response feedback loop that we use in order to deal with our existence in the world. But stop for a minute and think about what this actually means for language, then: that there really isn’t anything to it besides random associations. What to do with grammar, if there is nothing but random words? Well, we can also think of the word order that we use as nothing but a set of familiar patterns. Of course, I don’t believe that Bergen actually consciously comes to such a conclusion about language—namely, that language, grammar and syntax are essentially meaningless add-ons, and that other “more primitive” mental processes are actually what fill in the gaps between words that are more difficult to associate together—giving us only the illusion that grammar and syntax exist in the first place. For instance, the association between mouth and water, based on a spatial relation/stimulus-response pattern, is very easy to make. However, it is much more difficult for a person to make the spatial connection between a wagon pulled by oxen and a cliff, and the important detail that the wagon and all the oxen ran over the cliff, for the reason that the driver of said wagon had fallen asleep. Bergen may be unintentionally wrapping language up into a box of deconstructionist scrap for the sake of making the entire concept of language consistent with what is already known about how animal brains work. However, I don’t believe that Bergen’s research ultimately has the intent of totally deconstructing anything poetic about language itself. In fact, it is actually quite refreshing for me, as a traditional linguist, to see objective scientific research done showing just how utilitarian language actually is; most people don’t understand that language is actually a human being’s most important tool (I’ll come back to this topic in a bit).

            Don’t get me wrong, I strongly feel that this new linguistic concept of embodied simulation plays an extremely important role in the way that we as people use language in our everyday lives. Bergen points out relatively early on in his book an example of athletes who, instead of training harder in their sports by performing sports-specific tasks, actually made enormous improvements in their respective sports by merely imagining that they were performing, and acting out various athletic tasks in their heads (which efficiency was of course objectively measured with the help of brain scans and brain wave measurements). What this implies about language is, essentially, that it functions as a general “body command system” in the same way that actually performing various actions does. Furthermore, in this paradigm, there must be complete continuity between imagined action and the actual physical completion of an action. What Bergen has essentially tried to do is add language into this paradigm, and make no distinction between imagining an action, and actually constructing a coherent grammatical thought about performing the same action. Are these two actions the same thing? Can one imagine performing an action without also constructing a sentence? I don’t believe that Bergen ever asked this specific question in his book, but I believe that it is fairly obvious that visualizing and/or simulating something in one’s mind does not require the mental construction of a grammatical sentence. It is only when one is trying to make sense of the objects that one visualizes that one is required to use language, and it is my contention that  It may indeed be necessary to mentally construct a grammatical sentence in one’s head if one is to perform, or even understand, a specific action, but for the process of imagining any random object or a place, it should be fairly obvious that senses such as vision and hearing are perfectly adequate at helping to accomplish such tasks.

            What Bergen does not bring up in his book, which, in my opinion, really needed to be brought up, was the actual role that language plays in our lives. I believe that we can, without providing any statistics, experiments or empirical evidence, say with confidence that language regulates every and any human activity that can be undertaken—from complex discussions about abstract topics to taxing physical actions. It is the “master controller” of all of our activities. Think about the vast array of activities that it controls, and you will then realize that you would not be able to do anything without it. For a person to perform any sort of complex task, he or she must use language in order to understand the components of the task. Let’s take the example of brewing a sort of primitive alcoholic drink, or building basic tools for survival. Let us presume that there is a group of early humans who are looking to have a party; in the typical anthropological understanding that most people have of early humans, there is a lot of grunting, and not a lot of thinking—so how do they undertake the brewing process? What most people don’t realize is that early humans were every bit as communicative as we are today; the stereotypical “Tarzan” speech that lacks grammar is not sufficient for the needs of our survival in “the wild”. Early humans needed to be able to use grammatical language in order to explain the complex and often confusing set of circumstances that life as a hunter-gatherer would naturally demand in those days. And if you really think about it carefully, how could basic survival skills ever be accomplished if not for the ability to make grammatical sentences in one’s mind? For the sake of this “simulation” about life as a primitive human, let us imagine ourselves trying to accomplish any sort of task that requires planning without the ability to form sentences. This doesn’t mean just the lack of ability to utter sentences aloud, but to form mental sentences in our heads as well. Imagine your mind just seeing images of objects and living things, but not being able to tell yourself what you are going to do next in relation to them. Without grammatical commands in your head, you are totally unable to figure out what you are going to do next. So not only are we as human beings absolutely dependent on grammatical language in order to accomplish and achieve what we need to in our daily lives, but our brains themselves seem also to be exclusively dependent our ability to use grammar—that is to say, without grammar, we absolutely cannot function; or rather, if we are “planning animals”, then our brains are hard-wired to give us the ability to make those plans and not be bound to a set of “animal instincts”. If it is indeed the ability to make tools and plan complex scenarios that make us uniquely human as creatures, then the logical conclusion should therefore be that it is language itself that makes us human. Language is the most important tool that we as people have; language is the foundational tool. It is not just the babbling of primitive words that primitive humans used in order to name and list objects that gave rise to human language; rather, human language needs to have been fully grammatically functional since the time of its inception. Human language is not merely a set of Pavlov-ian cues that give rise to muscular responses; rather, it is the means by which human beings have the free will to set out to accomplish any and every thing that they could ever hope to achieve in the world.

            Of course, making such a statement about language on my part, since I am not a researcher with access to brain scanning equipment (and frankly, I have no real desire to do physical brain research), extremely subjective. I am relying entirely on my own intuition in order to make such assertions about human intelligence. And the reason I am able to do so is simply because the concept of “meaning” does not, and cannot, be broken down into anything smaller than it already is. Ultimately, understanding, or meaning, is the essence of our humanity—take that away from us through means of “objective research”, and we as beings don’t make a bit of sense in the context of the natural world. Cheetahs run quickly across the plains, frogs hop through swamps from lily pad to lily pad, and we humans reason in order to solve problems. There really is nothing more to say about any of it. What Professor Bergen seems to be implying with his book is that there is a great new scientific theory that is likely to arise out of scanning brains; that somehow, every neuron can be mapped in order to figure out exactly how we as people undertake our daily existence; there will turn out to be a typing neuron and a coffee neuron, etc..., and some day, we will have everything figured out in advance so that we don’t actually have to live our lives with any amount of uncertainty. But what scanning brains and deconstructing language to a set of neuron cues and pathways fails to do is take the concept of “free will” into account. It is such a simpler conclusion to arrive at—that the human experience is extremely subjective, and that we have freedom to choose our own actions. Early humans most assuredly never questioned their own existence, and consequently never labeled themselves as “primitive” or in need of a more advanced or objective consciousness; they were probably too frightened by their surroundings and were just trying to survive. For that reason, they most assuredly cherished their tribal languages and told vivid, poetic stories and sang songs in them from generation to generation. The final conclusion that any person dedicated to understanding human language must make is that language is goal-oriented; the goal of language is to produce meaning and understanding. I find the entire notion of deconstructing meaning to be, quite frankly, appalling, and though I do find it extremely interesting that we can scan our brains and realize that our entire bodies are involved in the actual understanding and producing of language, I nevertheless think that the quest to find specific “language neurons” to be completely absurd. As a traditional linguist, I refuse to deconstruct language simply because I do not want to deconstruct my own humanity.