Thesis continues:



When discussing the brain, neuroscientists modulate between expertise and simplicity. Their own topic of study-be it a neural nucleus, the biochemistry of bipolar disorder, the blood flow in brain structures during social behavior-is subject to rigorous and rich amounts of detail. Yet the areas on which the neuroscientist's expertise fails to cast light are numerous and present many interesting implications. I asked each of my participants about how they conceive of and think about the brain as a whole. Their responses were divergent. Some spoke freely about the brain, willingly exploring their conceptions of its totality without reservation. These researchers present a holistic approach. Others were careful not to stray far from their expertise. These individuals, the atomistic researchers, believe that moving too far away from their objects of research decreases their capacity for making meaningful statements. Yet, in their responses, both groups employed a nearly identical array of conceptual metaphors, consistently making sense of the concept of the whole brain in terms of other concepts. The conceptual metaphors referenced most often (and that form the basis for discussion in this section) include: brain-function localization and behavior, the brain as computational circuits, the integrated and irreducibly complex brain, and the brain as a mirror. The following analysis will explore the similarities and differences in the use and understanding of these metaphors in order to isolate certain implicit beliefs used in the neurosciences.
The constituent members of the atomistic and holistic research styles and their areas of study are listed below:
Atomistic
Aaron (Neurobiology of circadian rhythms)
Lauren (Brain imaging of bipolar disorder)
Holistic
Ian (Social neuroscience of autism and creativity; mirror neurons)
Albert (Mathematical modeling of brain development in songbirds)
Bill (Social cognitive neuroscience of subjective experience and reasoning)
Karen (Brain imaging of language development in humans)

Aaron represents the atomistic approach to research. When asked to discuss the whole brain concept he is skeptical of making far-reaching and unspecified claims, as is evidenced by this statement:

Aaron: If you poked somebody's orbital frontal cortex they might get moody or something like that, which, you know is kind of true, but it's like if you poked the 405 freeway you'd find that maybe the 10 would be free, but it's not because you poked the 405...you can't just say "this area is for that"...but some areas are, if you want to take that approach and you want to study something in great detail, you can say "this nucleus, in isolation from everything else, can act this way sometimes"...instead of attributing to the whole brain. To know the difference makes it so I'm pretty careful to say that.

In this excerpt Aaron touches on several dominant metaphors that are present in the field of neuroscience. First, to support his skepticism toward talking about the whole brain, he critically employs the "poke-effect" metaphor. The "poke-effect" metaphor combines several historical trends within brain research: phrenology, brain imaging, lesions, and neurosurgical research. Parallels have been drawn between the theory of phrenology of the eighteenth-century and modern brain imaging (PET, EEG, fMRI, etc.) in the their mapping of theorized faculties onto specific brain structures (Dumit 23). Likewise, brain lesions and accidental traumas have been used by humans for at least the past several centuries to analyze the effects of rough brain trauma on behavior. An example of this is ubiquitously taught in introductory-level neuroscience classes through the story of Phineas Gage. While Gage was working as a railroad foreman an explosion sent a tamping iron through his frontal lobes, forever changing his personality and behavior (Kandel et. al. 353). Much has been made of this common anecdote within the brain sciences, the most notable idea being the brain's ability to survive and be affected by harmful environmental stimuli. This stimuli can now be greatly refined by the practice of invasive neurosurgery, allowing surgeons to actually "poke" the brain and, as based on phrenological and image-faculty theory, perceive its effects.
Aaron derides any theory in which one can say "this area is for that", opting for a cautious and irreducibly integrated view of the brain. He claims that the best one can do in explaining the brain is to say that "this nucleus, when isolated from everything else, can do this sometimes". This is further supported by findings from his own research in circadian rhythms. By culturing his nucleus of interest, the suprachiasmatic nucleus of the hypothalamus, he can can alter its function by controlling its chemical and neural environments.

Aaron: Well just in a single nucleus that it is so small I think this is really interesting. Since obviously this is connected to other areas of the brain this makes me think that "okay these things are probably going to act differently in isolation than with each other and, even in isolation, there's a complex makeup for each of these different areas that are gonna function more like a circuit because you can take a single neuron out of anything and it's going to behave differently than it does in a population of other neurons and it's not just because one neuron is the same in every given environment and we don't know why. I don't know why.

By studying and understanding this tendency of his research object, Aaron has become weary of making claims about any emergent neural properties since, in varying environments, neurons can have divergent functions. He uses his personal visual and cognitive experiences with his nucleus to construct a theoretical stance toward the whole brain.
I found that all of my participants tend to adopt this approach. When asked about the brain in its totality, neuroscientists pull from predominantly visual experiences to make sense of their areas of ignorance. This is a process that is similar to what Lakoff describes as image schemas, wherein a source image (e.g. neurons in a petri dish) is mapped onto a target image (e.g. the whole brain) (222). There are differences as to how often this is done. Some researchers consistently relate their work to their whole brain concept, allowing both to mold and make meaning out of each other while others, as evidenced by Aaron, constrain themselves to their neural expertise.
A response given by Albert, a birdsong researcher interested in language evolution, represents the more holistic view:

Albert: I think to be a working neuroscientist, to be able to do hypothesis driven experiments, to be able to isolate variables and be able to make claims based on evidence you have to be working under the assumption that, well, to be a working neuroscientist who is interested in cognitive things, things that have to do with the mind, whatever that means, you have to be working under the assumption that the brain is the mind. Otherwise what are you doing?

Brain scientists like Albert who are concerned with biological and cognitive aspects of the mind have increased in number in the past twenty to thirty years due to the surge in genetic and imaging techniques that allow access to theoretical mental states, no matter how rough and intractable. Other researchers, such as those involved in behaviorist, cellular, biochemical, and biophysical analysis, approach the whole brain problem differently. They reflect a theoretical stance similar to that of the working neuroscientist presented in Patricia Churchland's now classic work Neurophilosophy (1986). These researchers believe, in one way or another, that: (1)"The time for theories has not yet arrived, since not enough is known about the structural detail," (2) "What is available by the way of theory is too abstract, is untestable, and is anyhow irrelevant to experimental neuroscience," (3) "You cannot get a grant for that sort of monkey-business". Furthermore, it often happens that a piece of research is undertaken, not in virtue of a larger program but instead because the researcher has mastered a certain technique, and there are always more measurements he can make (Churchland 403-404). Yet now, more than twenty years after Churchland's work, even these reductionist hard-liners are coming around to the cognitive approach, albeit often in non-explicit ways. Aaron's work is grounded in the study of neuropeptides and metabolic function but, interestingly, his motivation to do this work is grounded by the brain's observable output:

Aaron: I don't think (neurobiological research) is going to explain human existence and I don't personally look for that out of studying the brain. I think studying the kidneys could be interesting too. I think studying the digestive system could be interesting too. I find a little bit more fun in the brain because there is a behavior associated with it that is really complex and it's always in our face.

A jump is made from studying an anonymous group of neurons that theoretically produce the complex behaviors that are common to and known by all humans. This is done implicitly by Aaron for, even though he does not search for direct, existential meaning from his work, he does recognize a difference between it and studying the kidneys or the digestive system. By knowing and watching the behavior of himself and others, Aaron provides context for his work. It is in this way that a group of neurons can become conceptualized with a name and a function (e.g. the superchiasmatic nucleus that controls circadian rhythm).
Behavior, then, becomes a form of the visual representations and image schemas used to conduct and understand neuroscientific research. Yet many researchers call into question the importance of the visual, sometimes vehemently. In her study of brain imaging Beaulieu (2002) explains that "researchers insist they do not know the brain by seeing it, by making its activity visible" and that "denials of the importance of imaging, in a setting where visual representations are abundant, leave the analyst perplexed-and, indeed, at the heart of the claims about the contributions of brain mapping is a paradox" (56).
New fields that represent this paradox of the image are continuously being created within the brain sciences. Many seek to research aspects of mind and cognition by linking together cutting edge technologies with extant theories and methodologies. One such field that has gained attention, both in academic and lay circles, is social cognitive neuroscience (SCN). SCN combines the methodology of cognitive neuroscience with questions and theories from social scientific fields such as social psychology, economics, and political science (Lieberman 260). Due to its integration with functional neuroimaging, an infant scientific technology itself, the field has grown rapidly in the past five years. Its findings are diverse and hotly debated. The predominant imaging technology used in the field, functional magnetic resonance imaging (fMRI), has allowed experiment, data, and theory to be pushed to novel and unprecedented levels. FMRI allows researchers to locate brain functions and signals in three dimensions by studying the increased blood flow in neural structures (Kandel et. al 370). The logic says that the more blood a structure (e.g. auditory cortex) requires during a certain cognitive or behavioral task (e.g. listening to a Bach fugue), the more important that structure is to that task.
The use of fMRI in experiment design and data interpretation is a subject of ambivalence in the neurosciences. As Bill, a social cognitive neuroscientist, points out, when strongly controlled, fMRI provides the possibility for powerful correlations:

Bill: So it works out really well like when you analyze the brain data like sometimes you have things that are kind of like manipulation checks like when people are listening to things when you model it you want to show that your subjects are actually showing activity in their auditory cortex. If they are not then potentially there is something wrong with the auditory stimulus or potentially the subject wasn't listening because it is extremely reliable. Like in a study I am doing now 9 out of 10 subjects show activity in the bilateral auditory cortex and it's just nice that you did something psychological which is they are listening to sound but its very basic and nice that you have this really really reliable prediction about which pieces of the cortex are going to be lighting up in your analysis for that particular condition.

Bill perceives a capacity for fMRI to check itself and to provide controls. By introducing consistent and simple stimuli such as auditory signals, researchers can employ fMRI's imaging capacities to become convinced that their experimental environments (the fMRI machine, their cognitive and behavioral tasks, and the brain of the subject) are all static and in accordance with their notion of psychological reality. Bill recognizes the necessity of statistics in determining this. In his study "9 our of 10 subjects show activity" wherein the one subject who does not may simply not be listening. By controlling the fMRI environment and setting the experimental expectations, Bill is able to conclude that the statistical outliers must be involved in other non-related psychological tasks. Furthermore, Bill is not only able to conclude that their minds are elsewhere, he reserves the capacity to know which tasks the subject must be involved in. Bill's logic is backed up by a common sense understanding of psychological function: if a subject is not listening to what they are intended to, they must be listening to something else. It is a streamlined and simple approach to the mind, utilizing that most common form of reasoning: "if not A then B".
Yet the object that researchers like Bill are attempting to map the mind onto, the brain, is not seen by all to be streamlined and simple. Many neuroscientists, those working in SCN included, simultaneously recognize the power of fMRI while doubting certain aspects of its theoretical assumptions. Ian, for example, challenges the efficacy of comparative neuroimaging:

Ian: The experiment begins with thinking about behavior and then the mind…because the brain does one thing but the effect that that gives rise to in the mind is completely different. I don’t know, I feel like, and I don’t know if we know the answer to this yet, but I feel like similar brain activation patterns, and I’m not talking at the level of the cell because there need to be different firing patterns for different thoughts to come your mind, but at the level of the gross patterns we see in activation in terms of fMRI there’s probably very similar types of activation so you can look at two brains and they look lit up in very similar ways but the two people are doing completely different things.

Ian's conception of functional imaging allows for more theoretical uncertainty. He imagines a scenario in which the images of brain function may be identical while the behaviors and thoughts represented in those images may be widely divergent. Other researchers and authors have recognized this in different fashions. Steven Rose, for example, approaches this conundrum in terms of the individual. Rose writes that "imagers using PET and MRI have been able to develop algorithms by which they can transform and project the image derived from any individual into a 'standard' brain" and that "brains are so finely tuned to function, so limited by constraints, that anything more than relatively minor variation is simply lethal" (59). The "algorithms" Rose mentions are known as "average-brain models", computational devices used by neuroimagers to represent statistically normal brains, upon which research data can reflected and analyzed (Vul et al. 6). Average-brain models are products of technical convention wherein some models, the Talairach for example, have been used for over twenty years, dating back to the days before modern functional imaging. The idea being hinted at by both Ian and Rose, then, is that neuroimagers, in important ways, construct the brain and mind to fit their own expectations and assumptions.
Critics of SCN have argued along similar lines. A now famous 2008 paper by Vul et al. originally titled "Voodoo Correlations in Social Neuroscience" challenges the statistical and experimental designs used by SCN researchers. The authors argue that the reliability scores typically found in both psychological and neuroimaging tests are lower than certain outstanding scores found in SCN studies. SCN studies should, hypothetically, see scores that go no higher than the average of those of neuroimaging and psychology (around .74 out of 1.0 ) since methodologically SCN combines both approaches (4). Yet, time and again, SCN studies show scores well above .8 and higher, making some researchers, those in SCN included, skeptical of the claims coming from the field. Interestingly, while representing this skeptical camp in some ways, Ian is open to the holistic approach popular in fields such as this:

Ian: So I try to think of the whole brain, you know what I mean? Rather then thinking about necessarily when I design studies I don’t think about the brain so much as I think about myself and behavior. Because, I mentioned that a lot of my studies have to do with real life situations and I’m really interested in tangible things that people can relate to, I just think about my behavior.

Later in the interview he goes on to say that:

Ian: That’s one way to do it and it’s the not the most traditional way to do it nor the most methodical, you can see I do work in a lot of different areas so it’s a little bit like...what’s the expression…jack of all trades master of none kind of thing. But I am just fascinated by human behavior and so I think about it from that aspect of it first and then bring it back to the brain.

For researchers like Ian, and in certain ways for all neuroscientists, imaging is a simultaneously useful and problematic tool. It is challenging to read scientific meaning into an image, and even more difficult establishing a coherent, peer-accepted place for it within an established disciplinary paradigm. Furthermore, neuroscientists are notoriously ambivalent in their relationships with images, often attempting to fully degrade their use in formal talk and discussion. Yet, importantly, neuroscientists (along with other scientists and lay people alike) recognize and practice an innate ability to read cultural and social meaning into images. De Rijcke and Beaulieu (2007) explore the relationship between the scientific and cultural production of images. They write that once images are let loose into society, the "context of the subject's body, of the lab and of [the] high-tech expertise is removed and any concrete sense" and that "it is at best evoked, and what remains are rainbow colors superimposed on a floating brain" (737).
Conceptual metaphors are mapped onto visuals provided by imaging technologies such as fMRI and PET. The dominant image schema referenced by the participants is that of the computer, more specifically its constituent circuits and their characteristics. Turkle (1997) points out that the field of Artificial Intelligence has provided Western scientific culture with two major views of the computer: the computer as information-processor and the computer as an emergent intelligence (1094). While some of the participants spoke about ideas similar to the emergence concept, all of them touched upon the information-processing model where information is processed in a linear, input-memory-output system. This simpler model of modeling brain function, contingent upon the idea of interacting circuits, allows researchers to theoretically link various brain structures with functions, drawing lines between the groupings of colors found in brain images. To demonstrate the universality of the computer circuitry metaphor I provide the following quotes, one taken from each of the six interviews:

Aaron: I have also worked in an area where you have looked at it as a whole functioning working together brain and I think it's kind of a difference between looking at a macrosystems level of analysis like is this a major computer we are dealing with or is this just a specific circuit in the computer

Karen: ...all the neural circuitry that is involved with emotion regulation and those kinds of systems are able to wire up for maximum efficiency.

Bill: [Explaining motor control] Something like "when you pick up this spoon here, it seems very easy but there is all this complex crap going on in your motor cortex on the left side of your brain which sends signals down which eventually cross in your spinal cord which then go down fibers all the way to your fingers and contract the muscles around there but do so in very coordinated ways".
Ian: Stuff like that is really neat when you think about the brain, the way sounds come in, sights come in, and tastes and the front of the brain makes the decision of how to react to them or not react and then the motor system either executes or not.

Lauren: I suppose I see it as more of a computational machine that interacts with, takes in input from the environment and, trying to understand just the operating principles of the machine is the goal of many of the studies that I would be interested in doing.

Albert: So at a more abstract level you can conceive of each neuron as just a computational unit and you can conceive of the brain as just a vast network of nodes which is performing some computation and those nodes are arranged in groups and those groups are arranged in bigger groups and you have this hierarchical organization to a level that maybe other organisms don't have.

The participants speak of the brain in terms of hierarchies, circuits, computational units (neurons), wiring, and systems. These concepts are given agency through the participants' visual experience and knowledge. For example, the idea of a circuit, be it between electrical nodes or brain structures, is not a priori knowledge. An individual must have some form of practical or experiential relationship with a real circuit before they employ it conceptually. This is true for each of the constituent parts of the computer metaphor that were mentioned. I contend that this is due to the impact of imaging technology on neuroscientific knowledge and thinking, both expert and lay.
As mentioned before, decoding a brain image requires certain cognitive feats, some taken from cultural-at-large and some from statistical science. The ultimate goal in the decoding process is to create a holistic view of the image, to piece together the seeming chaos found in the pixels into a coherent and, hopefully, meaningful piece of data. Achieving this requires that each brain structure present in the image is provided a form of function, whether active or inactive, and a means of a connecting up with other structures. In response to this need a conceptual tool kit has been amassed. At its core is the concept of the circuited, information-processing brain. The brains circuits can be removed, as Aaron points out, and rewired in new ways. They can be manipulated by using pharmacological intervention, sensory input, and behavioral changes (as represented by the work of Karen, Bill, and Albert). Yet, in all cases, the circuits can only function and be studied if they are part of visualizable, integrated networks. The whole brain, in its modern form, only fully comes into being when visually captured at work, its highest forms of function caught in colors and pixels. It is in this sense that the increasing presence of brain images and their interpretations have influenced what, as part of the contemporary epistemological paradigm, neuroscientists can reasonably expect from their work and how they can understand it through language.
The discovery, and subsequent popularization, of mirror neurons has played a pivotal role in the construction of the dominant whole brain metaphors used today and are a prime example of how neuroscientists integrate structures and function in the brain. Through their investigation of macaque monkeys in the 1980's and 90's, Giacomo Rizzolatti and his team found that there are premotor neurons that fire when a monkey performs object-directed actions such as grasping, tearing, manipulating, holding, along with when the animal observes somebody else, either a conspecific or a human experimenter, performing the same class of actions (Iacoboni 529). These findings show that the cognition of the actions and intentions of others play a fundamental role in primate sociality. Making mirror neurons even more enticing is the fact that they demonstrably integrate the function of the premotor cortex, the inferior parietal cortex (a major area of sensory processing), and various aspects of cognition.
Neuroscientists, as evidenced by the response of my participants, construct various forms of meaning based on the existence of mirror neurons. For example, Albert perceives their role to be important to bird vocalization and, by extension, language.

Albert: So our birds actually have mirror neurons in their song system, mirror neurons seem to be, they are definitely in the parts of the brain that are important for language. Social interaction, imitation are definitely very important for language. One of the projects that I worked on as an undergrad was tracking infants and when they developed the ability to share attention with someone else with is kind of a pre-requisite for learning language.

Ian, a mirror neuron researcher himself, sees their existence as a defining feature of the human species and its connections with the external world.

Ian: Especially with the mirror neurons for example and the interconnectedness of everything and just that whole principle of everything being connected is so pervasive in yogic philosophy and now we know that maybe the neural basis of that is mirror neurons. And the reason we have such intimate connection with other people and less connection with a chimp, even though we do have a connection, and less of a connection with a chicken, and less with a cockroach is because these neurons modulate your firing depending on what is attracting them.

Popular pieces of research, such as mirror neurons, are theoretical open game for scientists and non-scientists alike. The most salient forms of popular research are those that integrate pre-existing fields of study, or at least show the potential to do so. Whether or not mirror neurons will be found to play a role in language evolution or confirm basic tenets of yogic philosophy does not matter as much as the belief that such findings are in fact possible. In this sense mirror neurons are a profound example of the ways in which neuroscientists create and use meanings and metaphors from the brain. They provide linguistic concepts that can be used to discuss the whole brain, validate dominant research technologies such as fMRI imaging, richly connect structure with function, and, as we will see, inform the nature of cultural identity and experience in important ways.