Minding the Brain more

Journal of Adolescent & Adult Literacy 54(5) February 2011 doi:10.1598/JA AL.54.5.1 © 2011 International Reading Association (pp. 316 –321) C O M M E N T A R Y Minding the Brain George G. Hruby 316 aving done some reading in the literature over the past 12 years, I would like to report: neuroscience research hasn’t proven very much yet about reading instruction. (Brain research on reading processes and development is more promising, though findings even there are far from definitive.) But brain research has proven fascinating to many educators nonetheless, and increasingly it has been making appearances, often trivial but occasionally profound, in the literature related to literacy instruction and development. We might expect to see continuing interest in the future. Yet, across the wider literacy community, it is those involved with adolescent literacy who seem least prepared to make sensible use of insights from brain research or to address in an informed manner the practical and theoretical challenges it might pose. We do not seem much interested. I find this unfortunate. The appeal of neuroscience to growing numbers of educators is not surprising; most people assume, correctly, that our nervous system is a crucially important locus for our learning and skill development, and many would go so far as to suggest that our brain is, in fact, who we are (a philosophically dubious proposition). Unfortunately, as Zambo (2008) has found, there is an inverse relationship between degree of conviction that neuroscience is of great significance for understanding learning and development and critical sophistication about such claims. (Indeed, in a more recent survey, Zambo and Zambo [2010] noted that education students who were most certain that neuroscience is important for educational practice were also those most likely to credit talk-show gurus Oprah and Dr. Phil as their prime sources of information on the topic.) The neurosciences are a specialized branch of the life sciences. Biology has long provided a theoretical foundation for developmental psychology (now more commonly termed developmental science). And developmental psychology has been a foundation for educational psychology, learning theory, and educational practice since the pioneering work of the functional psychologists at the turn of the 20th century (e.g., Dewey, 1896; Huey, 1908/2009). H An obstacle to understanding neural processing may arise from the field’s co-opting of cognitive psychological methods and metaphors. The informationprocessing model of cognition, which underlies much work in psycholinguistics, arose at a time when the Minding the Brain In literacy education, neuroscience research has been alluded to increasingly in work meant to substantiate cognitive models of the reading process related to text decoding and its instruction. As a result, this material tends to focus on early reading or reading disability and dyslexia (e.g., Hudson, High, & Al Otaiba, 2007; Shaywitz, 2003). It has also cropped up in educational psychology and special education texts for the same purposes (e.g., Alexander, 2005). And it can be found in abundant, if degraded, degree in the so-called “brain-based” and “brain-compatible” teacher materials that weave together loose reference to the brain, cognitive theory, and educational practice, apparently in the belief that juxtaposition and hand-waving assertion can happily stand in for cited evidence and logical warrant. This presents a curious caricature of neuroscience, although one in harmony with motifs found in the popular media—for example, the commonly encountered brain-as-wet-computer image. Few neuroscience studies actually target cognitive processes, and of those that do, few attempt anything more than to search for the neurological correlates to putatively cognitive behaviors. Few studies presume to demonstrate anything about the underlying theoretical assumptions of cognitive psychology, let alone cognitive educational learning theory. When theories of learning are in fact suggested (by neuroscientists, as opposed to cognitive psychologists in collaborations), they are often situated, dynamic, or bio-ecological in nature, but they are typically constrained to a biochemical or cytological scale. Only in education circles (and the popular media) does the belief go unchallenged that neuroscience is an obvious (and authoritative) handmaiden of cognitive educational psychology and its theoretical presumptions. By contrast, within neuroscience discourses—as encountered in scholarly journals or annual scholarly meetings—this bias is not in play and is occasionally refuted outright. As noted by Gernsbacher and Kaschak (2003), neural operations involved in cognition could mainly be discussed metaphorically.... [T]here is reason to suspect that nature has not cooperated by designing the brain to match our information processing intuitions. This likely explains why we observe the many-tomany mappings of structures and putative processes across imaging studies. (pp. 109–110) Few studies presume to demonstrate anything about the underlying theoretical assumptions of cognitive psychology, let alone cognitive educational learning I make this point to reassure theory those who are averse to cognitive motifs guiding literacy education. I believe the presumed connection to cognitive psychology, given its mixed history in facilitating improved literacy in classrooms and particularly its role in the so-called “reading wars,” may be one reason the prospect of an educational neuroscience leaves many professionals in adolescent literacy cold. But, as I say, this too-tight association between the two disciplines is a misrepresentation of the neurosciences, and it ill presages their potential to have a profound impact on education in the future. Why the Disaffection With Neuroscience? Content area reading, an important predecessor to today’s broader adolescent literacy, emerged during the 1970s between the twilight of behaviorist psychology’s positivism and the dawn of theoretically driven cognitive research, with its mechanisticcomputational orientation, a motif still heartily with us. This was followed by more contextually acute sociocultural literacy research, beginning from the late 1980s. A paradigm war of sorts was even brief ly engaged. Making a coherent connection between adolescent literacy’s two mainstream ontological metaphors—to borrow from Stephen Pepper (1942), mechanism and contextualism—and an organicist framework, grounded in the bio-ecological dynamics of neural development, seems an improbable endeavor. For one thing, even the more general domain of developmental science (e.g., Damon & Lerner, 2006) seems poorly understood in our corner of literacy. Only rarely does developmental process appear in 317 318 publications in our field, and when it does, it is likely to be grounded in the work of Vygotsky or Piaget, as if nothing conceptually significant has happened in the developmental sciences in 70 years. Turn the tables and imagine the absurdity of today’s developmental researchers studying reading by the light of Edmund Burke Huey or William S. Grey, and you begin to understand the reluctance of some psychologists to participate in educationally relevant transdisciplinary collaborations. If we hold but a weak grasp of current developmental science generally, how can we ever get a handle on the particulars of neuroscience? But perhaps adolescent literacy professionals face an even more daunting barrier in appreciating educational neuroscience. The aforementioned sociocultural turn in literacy education coincided with the coming of age of adolescent literacy as a field in the 1990s. Both veins of work were very inf luenced, both demographically and theoretically, by feminist and critical theorists who held an understandable aversion to biological idioms that had been used historically to subjugate less “advanced” or “evolved” peoples and populations (Lesko, 2001). In the late 19th and early 20th centuries, Galton’s phrenological analyses and emerging notions of social engineering potentially empowered by biological science led to essentialist arguments for the inherent inferiority of women and nonwhite races, and even, in some quarters, eugenics (Gillham, 2001). I do not find it unreasonable to fear that popular (as opposed to professional) interest in current neuroscience, genetics, and human development might suffer from some of the same taint. The media still lazily fall back on simplistic analogies, such as genomes as blueprints and neural networks as hard-wired circuit boards, even though the science clearly indicates these are wildly incorrect. Apparently, being wrong about the science does nothing to encumber circulation rates or audience share. Such static and essentialist idioms appeal to particular ideological orientations. All of these traditions, worries, and misperceptions are in play when adolescent literacy professionals try to get their brains around the brain (cf. Strauss, 2005). So allow me to demonstrate on a more manageable scale just one example of our ability to glean inspirational gems from the neurosciences. The Logic of Methodology and Data Interpretation: Brains and Achievement Gaps Well-replicated and explicated findings from other fields are the vaunted stuff of transdisciplinary synergy. But so can lapses in judgment be. Meet the methodological nugget called the nonindependence error from the critical neuroscience literatures, an authoritative hiccup that might prove useful in our thinking about literacy achievement gaps across grade levels and into adulthood. It has been lamented before that brain imaging is too often mistaken for photographs of the brain in action (Hruby & Hynd, 2006), when in fact they are statistical charts of subtracted correlations. Apparently, in constructing their data representations, even neuroscientists can be seduced by the allure of a good “image,” but, as I will point out, we literacy folk, too, fall afoul of the same seduction with our own statistical images. First, bear with me as I delve into some technical detail. As impressive as recent advances in the neurosciences may seem, work in this area is still largely formative and under a great deal of ongoing conceptual and methodological reformulation. Event-related potential studies began to emerge only in the late 1980s, and hemodynamic studies in the early 1990s. Because the number and diversity of these studies have expanded exponentially since then, the majority of imaging studies is less than a decade old and are usually unreplicated. Meta-analyses are scarce, and the few that exist are limited in scope. Taking claims at face value based on a sampling of such studies is ill advised (Hruby, 2009). Because of the novelty of the technologies employed, many brain imaging studies are, at least implicitly, as much about study design and methodology as they are about the neurological basis of the behavior under investigation. This makes for some very interesting and spirited debates in the neuroscience literature. Yet most of the critique revolves around conceptual fundamentals about which well-prepared literacy scholars are already knowledgeable: the difference, for instance, between necessary and sufficient conditions, between correlation and causation, and between reliability and validity. All have made Journal of Adolescent & Adult Literacy 54(5) February 2011 appearances in the history of both critical neuroscience and reading research. A case in point: Vul, Harris, Winkielman, and Pashler (2009) observed an abundance of “puzzlingly high” (beyond the likely upper bound) correlations in brain imaging studies on emotion, personality, and social cognition, given the presumed reliability of the imaging technologies and psychometric inventories employed in the research. Analyzing 55 recent studies, they found that the profoundly high correlations of localized brain activity to target behaviors were actually often due to “double-dipping” of the data set. Specifically, they found that researchers had selected precisely those brain-scan voxels (the neuroimaging equivalent of pixels in a digital photograph) for analysis that demonstrated a significant difference beyond an informed but essentially arbitrary signal– noise threshold to calculate the correlation of neural activity to behavior. (Brain images are based on subtracting neural activity data between an experimental condition and a comparison condition, but this calculation must be computed at each of the approximately 40,000 voxels that collectively make up each brain “snapshot.” Sorting noise from significance with these novel techniques is very much a fine art.) To demonstrate the simplicity of this error when unwrapped from the complexity of brain imaging methodology, Vul and colleagues (2009) used a similar design to calculate the correlation between stock prices and daily temperatures at a weather station in Arkansas. From the 3,315 stocks listed on the New York Stock Exchange, they selected only those that exceeded a correlational threshold to changes in temperature during a 10-day period. They then performed an analysis on only those winning stocks, discovering the weather station’s temperature readings powerfully predicted stock f luctuations (r = –0.87!). As this demonstrates, neuroscientists are apparently doing more than emphasizing the data for clarity’s sake; they are often giving them an undue emphasis to lead interpretation down a beautifully lit cul-de-sac. Such poor reasoning in the design and interpretation of brain imaging studies, particularly when coupled with small subject sets and a lack of replication, should be sobering. Vul and colleagues (2009) referred to this circular approach to Things grow finding meaningful relationships as the “nonindependence error” (p. confused when we 279), a type of fallacious reasoning try to compare the that has a vexing history in psycholiteracy achievement metrics (Barrett, 2009; Cureton, 1950). Cherry-picking data pre- of groups that are cisely because of high correlation identified by their value to perform correlation anal- literacy achievement ysis will indeed generate impressively high correlations. Although Vul and colleagues chose to do this analysis on social and personality neurosciences because of the recent popular excitement generated by those fields, they concluded that this problematic logic likely besets much cognitive and behavioral neuroscience as well. The debate that has ensued from their work has been both entertaining and informative. (See some initial responses in the May 2009 issue of Perspectives on Psychological Science, the journal in which the Vul et al. article appears.) Something like this fractured reasoning crops up in our thinking about literacy achievement gaps, too. It is certainly reasonable to compare the literacy achievement of groups identified by nonliteracy factors (gender, socioeconomic status, race, etc.). But things grow confused when we try to compare the literacy achievement of groups that are identified by their literacy achievement. As with the nonindependence error, an informed but essentially arbitrary threshold of literacy achievement is set, below which a student is deemed inadequate in reading (“struggling,” “striving,” “disabled,” etc.). The scores that identify this subset are then double-dipped to compute a mean score for comparison with the mean of scores above the threshold. Obviously, there is going to be a difference between the two mean scores—the data have intentionally been categorized on just that basis—and this difference is quickly labeled a “gap.” When plotted over time, the gap increases. On the basis of this shocking (but inevitable) result, the warning is issued that if we do not rescue those individuals below the threshold early, their disabilities will grow worse. Actually, though, this “gap” is only a visualization of what had already been presumed true—that Minding the Brain 319 320 the threshold is a meaningful distinction, something we feel more certain of, in part because it is now graphically visible. Just as rainbow-colored fMRI charts are easily misinterpreted by neophytes as demonstrating modular neural activity dedicated to particular functions, or as electroencephalogram charts are misunderstood to suggest “brainwaves,” so, too, does the achievement gap become a reality through the miracle of visualized statistics. Were we to disaggregate the underlying data and look at the course of the individual trajectories that make it up, what we would see is no gap at all, but the typical distribution of variance around an overall population mean located between the subpopulation averages, a distribution that increases with age (and probably, when corrected for other variables, with a fair degree of boundary crossing). In other words, what we have is not really a gap, but an increasing achievement spread. Reducing this literacy achievement spread on behalf of more consistent literacy achievement in the general population may be a laudable—indeed, democratic—goal. But grouping and meaning the data this way obscures the underlying fact that we do not really have two distinct reader populations growing ever more distant (unless we choose to segregate them that way on the basis of an arbitrary dividing line). What we have is one population united in being increasingly variable over time (enter adolescence!). The juxtaposed similarity of this gap with the more validly demonstrated achievement gap of historically marginalized populations serves a rhetorical trick: struggling readers are, clearly, distinctly different from the rest of us and therefore ought to be segregated. It is not impossible that further marginalization of the historically marginalized is the subversive intent. So one of the first things neuroscience reminds us of is that we should be very cautious about statistics reified into images that are too easily mistaken to represent tangible, rather than constructed, aspects of reality. Thinking of developmental change inaccurately as a growing binary-oriented gap suggests a looming societal chasm (down which we might presume a terrifying abyss doth suck), and can incite panicked calls of “crisis,” “emergency,” and “systematic failure”— and a scrambling away of middle and professional class support for public education. There are other methodological issues that have been cause for concern in the neuroscience research literatures (e.g., Bennett & Miller, 2010; Bennett, Wolford, & Miller, 2009). Many of the topics should sound familiar to well-prepared education researchers. Distilled, they provide us with a double-barreled caution: 1. Nonspecialists outside of the neurosciences should be cautious about taking any particular brain study finding or claim at face value (particularly when disseminated through the popular media), let alone as a definitive form of evidence for a reading program, method, policy, or theory. 2. We should be on the lookout for circular reasoning on behalf of what we are determined to find in our own use of quantitative measures, and be more critical of the social policies that grow out of such misuse. Back to the Big Picture: Brain-Literate Brains of the Future As I have hinted throughout this commentary, the bio-ecological nature of the brain and its development within education contexts may not be well parsed by reductivist or computationalist theoretical frames for understanding thought and being. The insights bound to emerge—and soon—from the neurosciences of learning are, like the neurosciences themselves, essentially biological, or bio-ecological. Therein sits the promise and the challenge for adolescent literacy professionals trying to come to terms with it all. I believe, though, it is a nest of tensions worth delving into. The insights that are going to rock our world from the neurosciences (and the developmental sciences more generally) will not be about computer simulations of neural network models or the neophrenology of color-coded brain maps. The impact of what the neurosciences have to tell us about literacy development, language, and learning will arrive when biochemical and anatomical pathways have been mapped from environment to genome (and back again) indicating quite profoundly, as a scientific matter, that our environmental history is a major factor in the nature of our social, emotional, and intellectual Journal of Adolescent & Adult Literacy 54(5) February 2011 development and our educational propensities. This mapping will extend clear down to the genome and its expression through histone binding and methelation, to the protein chemistry generated thereby, and thence to the functional modifications within synapses, neurons, networks, and systems that mediate our learning and growth as social beings. It will provide the key, in other words, to the old nature–nurture controversy. And this will not be good news for ideologically committed nativists. When will this occur in earnest and reinvigorate discourse on developmental processes in education (and the need for greater political support of educational environments)? Soon. I will stick my neck out and predict that it will be obvious within 10 years. The genetic particulars of this insight are already coming into print. It is only a matter of time before research and theoretical implications extend into the developmental neurosciences, then to arenas of public policy debate, and from there to the classroom. We need literacy education professionals prepared for this, and prepared to inform and lead the policies and practices it will spawn. References Alexander, P.A. (2005). Psychology in learning and instruction. Columbus, OH: Prentice-Hall. Barrett, L.F. (2009). Understanding the mind by measuring the brain: Lessons from measuring behavior. Perspectives on Psychological Science, 4(3), 314–318. doi:10.1111/j.1745-6924.2009 .01131.x Bennett, C.M., & Miller, M.B. (2010). How reliable are the results from functional magnetic resonance imaging? Annals of the New York Academy of Sciences, 1191, 133–155. Bennett, C.M., Wolford, G.L., & Miller, M.B. (2009). The principled control of false positives in neuroimaging. Social Cognitive and Affective Neuroscience, 4(4), 417–422. doi:10.1093/ scan/nsp053 Cureton, E.E. (1950). Validit y, reliabilit y, and baloney. Educational and Psychological Measurement, 10(1), 94–96. doi:10 .1177/001316445001000107 Damon, W., & Lerner, R.M. (Eds.). (2006). Handbook of child psychology, volume 1: Theoretical models of human development (6th ed.). New York: Wiley. Dewey, J. (1896). The ref lex arc concept in psycholog y. Psychological Review, 3(4), 357–370. doi:10.1037/h0070405 Gernsbacher, M.A., & Kaschak, M.P. (2003). Neuroimaging studies of language production and comprehension. Annual Review of Psychology, 54, 91–114. doi:10.1146/annurev.psych.54.101601 .145128 Gillham, N.W. (2001). A life of Sir Francis Galton: From African exploration to the birth of eugenics. New York: Oxford Press. Hruby, G.G. (2009). Grounding reading comprehension in the neuroscience literatures. In S.E. Israel & G.G. Duffy (Eds.), Handbook of research on reading comprehension (pp. 189–223). New York: Routledge. Hruby, G.G., & Hynd, G.W. (2006). Decoding Shaywitz: The modular brain and its discontents. Reading Research Quarterly, 41(4), 544–556. Hudson, R.F., High, L., & Al Otaiba, S. (2007). Dyslexia and the brain: What does current research tell us? The Reading Teacher, 60(6), 506–515. doi:10.1598/RT.60.6.1 Huey, E.B. (2009). The psychology and pedagogy of reading. Newark, DE: International Reading Association. (Original work published 1908) Lesko, N. (2001). Act your age: A cultural construction of adolescence. New York: Routledge Falmer. Pepper, S. (1942). World hypotheses: A study in evidence. Berkeley: University of California Press. Shaywitz, S. (2003). Overcoming dyslexia: A new and complete science-based program for reading problems at any level. New York: Knopf. Strauss, S.L. (2005). The linguistics, neurology, and politics of phonics: Silent “e” speaks out. Mahwah, NJ: Erlbaum. Vul, E., Harris, C., Winkielman, P., & Pashler, H. (2009). Puzzlingly high correlations in f MRI studies of emotion, personality, and social cognition. Perspectives on Psychological Science, 4(3), 274–290. doi:10.1111/j.1745-6924.2009.01125.x Zambo, D. (2008). Childcare workers’ knowledge about the brain and developmentally appropriate practice. Early Childhood Education Journal, 35(6), 571–577. doi:10.1007/s10643 -007 -0223 -2 Zambo, D., & Zambo, R. (2010, May). Educator’s beliefs about neuroscience in education: Promises and concerns. Paper presented at the annual meeting of the American Educational Research Association, Denver, CO. Hruby is an associate research professor of literacy education at the College of Education, University of Kentucky, Lexington, USA, and the executive director of the Commonwealth of Kentucky; e-mail george.hruby@ uky.edu. Minding the Brain the Collaborative Center for Literacy Development for 321