Abstract
The mission statement of the American Educational Research Association (AERA) articulates the following goals: “improving the educational process,” “advancing knowledge about education,” “[encouraging] scholarly inquiry related to education,” and “[promoting] the use of research . . . to serve the public good.” I examine AERA presidential addresses from 1948 to 2010 as windows into how the research community has conceptualized learning and explore the following questions: (a) What has evolved in the science of learning? (b) With what persistent questions has the field wrestled? (c) In what ways has the field addressed broader factors—social, cultural, political, economic—that represented the ecological contexts in which education in formal and informal settings played out in each decade?
It is no surprise the science of learning has been largely situated in psychology, particularly, educational psychology. At the same time, there are periods in which interdisciplinary theoretical and methodological collaborations emerge across disciplines. From the 1940s, when this review begins, until now, the biggest shift has been from a focus on individual mental functioning to a broader focus on people’s participation in social and cultural practices. An additional lens has been to consider implications for development over time (e.g., child and adult learning within and across the life course; the development of expertise) as well as implications for learning across physical spaces. Finally, there have been persistent inquiries into cognitive functioning within broader neurological and physiological processes. I examine both shifts and continuities in light of Ann Brown’s warning in her 1994 address: “We repeatedly throw out babies along with bathwater, when we should build cumulatively. No community can afford to lose so many valuable offspring in the service of progress” (p. 2). I think of shifts as fundamental reconceptualizations and continuities as expanding the dimensions around which we seek to understand a persistent phenomenon.
Evolution of the Science of Learning: Shifts and Continuities
While I document these shifts through analyses of American Educational Research Association (AERA) presidential speeches from 1948 to 2010 that explicitly address issues of how people learn, I want to invite the reader to examine these shifts through a critical lens. Across these theoretical shifts remains a set of underlying dilemmas: Is learning simply internal to the individual or shaped by participation in practices? Is human learning characterized by finite capacities or plasticity? Are there underlying hierarchies or gradations in the quality of what we might call cultural practices that provide supports for learning? In the broad arena of education research on learning, these questions have been explicitly articulated and tackled, whether through a behaviorist lens, a cognitivist lens, a sociocultural lens, or an ecological lens. And within each of the historical time periods when these frames have been most dominant in theory building and practice, there has been research highly touted and valued that puts forward propositions about some sets of people, some sets of practices, some ways of using language, some ways of organizing families, and some psychosocial states that are inherently superior and predictive of academic achievement. And unfortunately, these deficit views focus largely on youth, adults, and families from nondominant, minoritized groups and groups living in poverty. I am not arguing that all practices and modes of reasoning lead to equal outcomes, but rather, we have legitimate differences around what outcomes count for what purposes and about the range of diverse pathways through which humans can accomplish even shared goals. So among the questions with which we must wrestle as researchers is why these themes of inherent deficits or fixable deficits persist, despite shifts in our theoretical frames. It should be noted that there have been persistent challenges to deficit theorizing across research fields. However, these bodies of research seldom become integrated into, for example, research reviews, widely funded research projects, or proceedings of various academies unless the topic is explicitly about diversity and equity along some dimension. As one example, from reading mainstream journals in the various fields of psychology, one would not even know that the field of black psychology even exists.
But let me be clear. I do not offer this critical lens as a critique of the individual addresses and certainly not of the individual AERA presidents, but rather, I seek to contextualize the issues about learning they address as sitting inside of broader intellectual and theoretical groundings in our related fields of study impacting education.
The Behaviorist Frame
Brownell’s (1948) speech addresses challenges to and limitations of the dominant behaviorist explanations of learning. Behaviorism dominating the first half of the 20th century focused on external behaviors as observable and measurable evidence of learning rather than on internal, nonobservable characteristics of mental functioning.
Although we may choose to think of behaviorism as an antiquated framework for understanding and examining human learning, in her 1994 address, Brown warns, “[Behaviorist] theories, albeit somewhat disguised, are still alive today” (p. 3). Although Brown asserts this as a critique, there are current new applications in the field of applied behavior analysis used in the field of special education (Mace, 1994). Brownell (1948) describes “[behaviorist] learning theories . . . [as] devised to account for what happens in learning under pre-arranged and controlled conditions” (p. 491). In 1994, Ann Brown raises a similar critique of studies of animal behavior from this same tradition presumed to extrapolate to human behavior: Some [toddlers] were asked to stack boxes or use sticks to obtain objects out of reach, just like Sultan the chimp. . . . It did not seem to occur to anyone that a set of boxes more readily affords climbing to an ape than to a less agile human toddler. (p. 2)
However, Brownell goes on to say, “The classroom situation consistently involves a multiplicity of inter-related S’s, many of them not known, to say nothing of their being controllable. . . . We cannot eliminate the complexities of the classroom by ignoring them” (Brownell, 1948, p. 487). In his 1965 address, Cronbach revisits Brownell’s dilemma as one deeply rooted in the broader field of psychology: Persons concerned with education found no nourishment in the systematic Hullian studies of T mazes and eyeblink conditioning that began to dominate experimental psychology in the late 1930’s [sic]. Experimental psychologists were repelled by the educator’s insistence on talking about “the whole child” in “real-life situations”—both being prescientific or even antiscientific phrases, antithetical to the analytic, formal style that, for a time was the ideal of American behavior theory. (p. 110)
Thus issues about the questions to be raised, the contexts under which to study, and the methods to be employed for studying learning were and indeed continue to be sites of disputation.
Moving From Behaviorism to the Cognitive Revolution
In many respects, from across the presidential addresses selected, Gagne (1972) represents an intellectual bridge from the behaviorist focus on influencing external, observable behaviors through stimulus response to the cognitive revolution focusing on internal cognitive processes. Gagne began his career engaged in experiments with animals but moved to thinking about the design of practical instruction (in schools and workplace settings, including the military). In his decades of research on instructional design, he sought to identify behavioral conditions that he hypothesized would stimulate cognitive processes. The behaviorist tradition hypothesized that external stimuli conditioned internal responses. The cognitive orientation hypothesized that an external stimulus conditioned a response but one mediated by internal cognitive mechanisms. Through such mechanisms, learners perceive what they think is the task, make sense of it, and make connections and transformations of understandings as they seek to solve problems. Gagne bridges these two orientations. In many ways his propositions about how to design instruction are not far from the stimulus response tradition and actually connect well with recent foci in K–12 teaching and curriculum on direct instruction: gain attention, inform objectives, present stimulus material, provide guidance, elicit responses through performance, provide feedback, assess, and enhance transfer. It is interesting to note that direct instruction is more often than not used in schools serving students living in poverty than in schools serving middle- and upper-middle-class students (Milner, 2014). On the other hand, his categories of kinds of learning tasks align with the focus in cognitive orientations to learning in terms of specifying the cognitive demands of different kinds of learning tasks (e.g., learning verbal information, engaging in intellectual skills and cognitive strategies, engaging in motor skills, and learning attitudes toward the learning goals or tasks). Gagne argued that there exists a hierarchy of prior knowledge that is required to learn higher-order skills and problem solving. Gagne’s taxonomy of different learning tasks differs, for example, from Bloom’s by incorporating attitude and motor skills.
A number of propositions put forth by Gagne (1972) continue to have traction. His work on instructional designs can be thought of as a foundation for what in the learning sciences is called design-based research (Barab & Squire, 2004) or what Brown in her 1994 address called design experiments. Current work on learning progressions (Heritage, 2008) within domains is likely influenced by Gagne. However, the focus in the current work tends to be discipline specific, and what is thought of as generative knowledge that may be transferable as students move on systematically to increasingly more complex concepts and kinds of problems is not limited to what Gagne saw as generalizable skills within a domain (e.g., progressions within remembering verbal information; or within intellectual skills, which he defined as the ability to distinguish among classes of objects and concepts by applying what we might think of as procedural rules; or within cognitive strategies, which he defined as largely metacognitive skills).
A major shift, however, in the evolution of cognitive frameworks for examining how and what people learn is a very different set of assumptions with regard to what novices need to know before they can learn more complex processes and problem solving. These distinctions are important and have had grave implications for children’s opportunity to learn, especially children from low-income communities. I’ll offer two examples. In early reading instruction, it has been presumed that if children do not enter school with knowledge of the alphabet and experiences with book reading, they are not ready to engage in comprehension. In fact, significant federal investments were made in diagnostic assessments, curriculum, and professional development to promote phonemic awareness because it was believed that insufficient phonemic awareness inhibited children’s abilities to comprehend texts. While there is no question that the development of phonemic competence is an essential component of reading, for young children learning to read and for older children who are poor readers, there are still multiple pathways through which children can be supported in learning to comprehend without withholding attention to comprehension until children master letter-sound relationships (Snow, Burns, & Griffin, 1998). The whole-language movement offered evidence that children’s sensemaking of visual information in children’s books as well as their linguistic repertoires, for example, around producing and comprehending a variety of genres of narratives from their everyday life experiences, as well as prior knowledge of topics about which they were reading, all contribute to the development of reading comprehension (Stahl & Miller, 1989). But underlying and deeply ingrained assumptions about hierarchies of learning continue to prove to be difficult conceptual hurdles. Such assumptions constrain the field’s abilities to think about how to expand the range of pathways through which children can learn. This is especially a conceptual challenge in the field with regard to children living in poverty and those with histories of underachievement as measured by existing tests used for accountability. In mathematics education, both the National Council of Teachers of Mathematics standards and the current Common Core State Standards do not assume, for example, that if children do not have sufficient skills in one particular arithmetic skill (e.g., not remembering one’s multiplication tables or division tables), this precludes them from learning what would be considered higher-order operations (Schoenfeld, 1987). There is also interesting, more recent work focusing on how children intuit fundamental concepts (and misconceptions) from their experiences in the natural world about number and physical causality, with the implications that such intuitive understandings or inferences can serve as scaffolds for formal learning (Gelman, 1990). There is also work on cultural models of relations in the natural world (e.g., Native American children either living on tribal lands or maintaining traditional relations with elders on tribal lands construct models of ecological relationships among living species, including humans, animals, plants, water, the land) that have been shown to be resources for concept development and intellective dispositions in science learning (Bang, Medin, & Altran, 2007).
Gagne (1972) argued that the categories of learning tasks were distinct, each requiring different instructional supports. However, more current research suggests that robust learning environments can and often entail all of the categories identified by Gagne but not as separate or distinct tasks. For example, there is an emerging field of embodied cognition (Varela, Thompson, & Rosch, 1991). Embodied cognition moves beyond the traditional focus in cognitive psychology on the brain as the seat of thinking and problem solving. Instead, the argument is that the entire body acting in the world and responding to the environment (physical, social, cultural) coordinates among its systems to make sense (all of the five senses and the pathways through which they coordinate with each other and with the brain and from the brain back to other physiological systems; Lakoff & Nunez, 2000). Lakoff and Johnson (1980) illustrate this process in what they call conceptual metaphors, where communities of humans construct linguistic metaphors that capture psychological inferences derived from bodily activity and observations in the natural world (e.g., moving up as positive psychological growth, as in having high standards). Embodied cognition is also being drawn upon in the design of instruction, especially in domains like mathematics and science. One example can be seen in the Algebra Project (Moses & Cobb, 2001), where children learn algebraic concepts, like reconceptualizing number as embodying not only quantity to now include directionality from a point of origin by taking trips on an urban transit system (an embodied experience) to capture their decision-making processes in mapping out transit routes to more distal representations through art, natural language, and then to mathematical representations.
Two presidential addresses after 1970 suggest an emerging shift from a primary cognitive orientation to a sociocultural focus. Glaser in 1972—a year after Gagne’s address—poses the question of how assessments and instruction can be responsive to individual differences that learners bring. Glaser argues that both uses of general IQ assessments and assessments of specialized aptitudes (e.g., spatial, mechanical, and abstract reasoning processes) are not strong predictors of academic achievement in and of themselves. He makes a distinction between what he calls “selective modes of education” that presume singular pathways for learning and are intended to weed out persons who do not measure up to the criteria for excellence presumed. He argues that uses of testing at that time were dominated by the selective focus, and I would argue that trend continues today, not only in testing for college admissions but increasingly for school-level and teacher-level accountability in K–12 public education, with the same detrimental effects that Glaser described in 1972. By contrast, he advocates for what he calls an “adaptive mode of education . . . that . . . can provide for a wide range and variety of instructional methods and opportunities for success” (Glaser, 1972, p. 6). Two points made by Glaser suggest a shift from a purely cognitive orientation to more of a sociocultural focus. First, he calls for attention in instructional design and assessment to what today would be called noncognitive factors (Farrington et al., 2012) that come into play in the repertoire of resources on which learners draw in acts of learning. In school intervention work, he designs for self-management skills, learning-to-learn skills, and for what he describes as “measures of process and style, of cognitive and non-cognitive development, and of performance in more natural settings than exist in the traditional school” (Glaser, 1972, p. 12) His is also the first in the series of talks I examined to explicitly address issues of cultural diversity as one source of understanding individual differences among students. While the language he uses is reflective of discourse at the time around cultural differences (e.g., “culturally disadvantaged children,” “cultural deprivation”), he is clear when he says, “In the adaptive educational environment that I envision, it would be assumed as a matter of course that the values, styles, and learning processes that the child brings to school are of intrinsic worth” (Glaser, 1972, p. 10). In addition, articulating a view of learning as situated and social, Glaser says, We now know that psychological functioning is a continuing reciprocal interaction between the behavior of an organism and the controlling conditions in the environment. . . . This is a two way causal process in which the environment might be just as influencable as the behavior it regulates. (Glaser, 1972, p. 11)
In a similar vein, Anderson in his 1984 address invokes both a cognitive focus on the role of schema in cognitive processing but also explores the ways that what he calls a “weak theory of schema” opens up a conceptual space for understanding the power of variation in schema, including cultural construals of phenomena from across different human communities. He says, “Weak schemas bring some order to messy words in which the sense and reference of terms is indefinite, categories are fuzzy, and the relationships among entities may be arbitrary” (Anderson, 1984, p. 8). Prior research on schema theory has demonstrated how readers’ interpretations of texts are influenced by the relevant schema they bring with regard to topics in expository texts and characters and themes in narrative texts, suggesting that meaning is not merely inherent in the text but an outgrowth of the interactions between what the reader brings and the text demands. Anderson and his colleagues at the Center for Reading at the University of Illinois–Urbana demonstrated how differences in cultural construals, for example, around what activities and goals are required for a wedding in India versus the United States, influenced very different inferences on a generic story about a wedding from readers from the two countries (Steffensen, Joag-Dev, & Anderson, 1979). In that work and in his presidential address, Anderson refers back to Bartlett’s work back in 1932 that introduced the idea of culturally specific schema. In the address, Anderson explicitly says, “Knowledge in turn is conditioned by culture” (Anderson, 1984, p. 8).
In the long-standing discourse about cultural differences in a wide array of practices, including presumed class-based differences, Glaser (1972) and Anderson (1984) weigh in with slightly different perspectives. Weigh in here may be too strong a term, as the references in both addresses are not the dominant focus of their arguments. Glaser makes reference to the idea that the middle class child acquires these things [particular cognitive styles] so that they are continuous with what will be required of him in school. . . . An obvious example in the conventional school is that, explicitly or implicitly, the school requires the immediate acceptance of an achievement ethic with deferred future rewards, a characteristic most consonant with middle class values. (Glaser, 1972, p. 10)
This line of reasoning is as strong today as it was both in the 1970s and long before (Jensen, 2009). It assumes the middle class and those living in poverty to be monolithic and assumes some conception of middle-class values is the social capital needed for productive learning in school. At one level this is true, because as Glaser points out, schools were then and are now organized around what he calls a “selective mode of education.” But at this time, the presidential addresses focusing on learning are not speaking in explicit ways about multiple pathways for learning. Anderson picks up on several of the early instantiations of the theory of cultural deprivation, particularly focused on language practices: There is a tradition of research that attempts to explain ethnic and social class differences in terms of “restricted” and “elaborated” linguistic codes. This research was based on the assumption, which I do not accept, that there are inherent differences in cognitive capacity. . . . The issue goes beyond discourse style. It is a question of assumptions about knowledge. (Anderson, 1984, p. 9)
The Sociocultural Shift
From the selected presidential addresses I have explored, I place Resnick’s (1987) presidential address as the first explicit shift to a sociocultural framework. I define here a sociocultural framework as focusing not solely on individual mental processing but the ways in which by interacting with other people, with the cognitive and physical artifacts available through human history, and with a focus on people’s participation with practices within and across spaces, people engage in problem solving and acts of learning (Cole, 1996). In the broader context, by 1987 there are extensive translations of Vygotsky (Wertsch, 1985) and an emerging body of work on learning in everyday contexts—workplace settings in the United States and abroad, domain-specific learning in nonschool environments—some of which Resnick cites, although she does not cite Vygotsky. She talks about distributed cognition and the function of tools and artifacts in problem solving. Resnick explicitly says, “Work, personal life, and recreation take place within social systems, and each person’s ability to function successfully depends on what others do and how several individuals’ mental and physical performances mesh” (Resnick, 1987, p. 2).
Specifically, Resnick (1987) distinguishes between the demands for participation in acts of learning in informal settings and those in traditional schooling. She does not make a simplistic argument that one is superior to the other but, rather, identifies the particular affordances and constraints in how learning is organized and toward what ends. She particularly emphasizes how problem solving is typically shared in informal and workplace settings versus demands for individual, often decontextualized performances in school. She says, Outside school, actions are intimately connected with objects and events; people often use the objects and events directly in their reasoning, without necessarily using symbols to represent them. School learning, by contrast, is mostly symbol based; indeed connections to events and objects symbolized are often lost. (Resnick, 1987, p. 3)
She wrestles with the question of under what conditions is learning most likely to transfer to other settings and related tasks and concludes that features of contextualized collaborative problem solving with objects characteristic of learning outside school along with the cultivation of symbolic reasoning contribute to generative learning. Resnick notes—reflecting on the challenges, then and now, of preparing people to participate in changing workforce demands—that we must “mount detailed examinations of people coping with situations of breakdown or transition in their work” (Resnick, 1987, p. 18).
As we see in Resnick’s (1987) address, this sociocultural shift encompassed increased attention to microgenetic processes entailed in how people learn in informal settings. Research in this area documented how youth and adults from nonprivileged and nondominant communities engaged in complex problem solving in a number of domains: Scribner’s (1984) study of workers in a dairy factory, Saxe’s (1988) study of Brazilian children selling candy on the streets, Lave’s (1977) study of Liberian tailors, and later, Nasir’s (2000, 2002) studies of high school basketball players and children learning to play dominoes. These studies documented strategic reasoning, including collaborative supports, and demonstrated conceptual links to problem solving in academic domains. Yet in the midst of this evidence, there have been few links to schooling, and there remained then and now persistent calls for scripted curriculum, basic skills, tracking, and research-based assertions reflecting deficit assumptions with regard to youth and families from nondominant groups living in poverty. This ongoing conundrum highlights even further the disconnect between theory and practice noted as early as Brownell’s (1948) address.
From Theory to Practice
Brownell (1948) criticizes Thorndike’s attempt to design instruction along behaviorist, stimulus response principles in mathematics back in 1924. Among his criticisms is the lack of sufficient detail to be useful and replicable in practice. He discusses the relative absence of meaningful explanation; little developmental instruction; a deliberate avoidance of most mathematical relationships which tie the myriad items of arithmetic into a unified whole; a corresponding emphasis upon the mastery of isolated items; and the tendency to give children at the outset the responses as they are eventually to be employed. (Brownell, 1948, p. 484)
These 1948 limitations of behaviorist designs for instruction are taken up proactively in Ann Brown’s (1994) articulation of design principles for the instructional framework Community of Learners (COL). In these design principles, she attempts to be explicit about underlying principles without specifying the kinds of instructional scripts so dominant in curriculum in the United States in the 1960s and today. In particular, Brown attempts in COL to address Brownell’s critique of what in 1948 he called “learning by doing” and what would eventually come to be known as “problem-based learning.” Brownell characterized learning by doing as trial and error, not providing sufficient guidance and scaffolding for novices. COL and other examples of cognitively guided instruction, reflecting both the cognitive shift and the sociocultural shift in dominant framing around how people learn, merge the use of explicit pedagogical scaffolds with what are often discipline-specific, cognitively focused articulations of
Across the selected presidential addresses, Brown delves most deeply in articulating what by 1994 served as a bridging of the cognitive and sociocultural frameworks for understanding learning. Issues of developmental readiness for particular kinds of learning goals had been an issue of concern across the decades. Brown merges these two traditions in articulating the following COL design principles:
“Academic learning is active, strategic, self-conscious, self-motivated and purposeful”
“Classrooms . . . [are] settings for multiple zones of proximal development”—creating opportunities for students with different beginning skill levels to have authentic pathways for engagement
“Legitimization of differences”
“A community of discourse”—creating practices that support dialogue and reciprocity as mechanisms that facilitate internalization of new ideas
“A community of practice”—designing activities that engage students in joint research and creating interdependencies among students as active members of an intellectual community
“The need for deep conceptual content that is sensitive to the developmental level of the students”
“The need for assessment procedures that are authentic, transparent, and aligned with the curriculum.” (Brown, 1994, pp. 9–10)
These design principles explicitly structure opportunities for collaboration as students research specialized aspects of a broader and conceptually complex problem, providing an authentic context and opportunity for all students to contribute to the overall understandings of the group. Brown invokes Vygotsky’s (1978) construct of a zone of proximal development (e.g., the idea that learning targets should be just beyond what the learner can do on his or her own). It is also interesting that Brown’s examples of learning targets are the most conceptually rich of any offered in previous presidential addresses, I think in part because COL was very focused on helping children explore complex constructs and systemic relationships within domains, especially science. Mapping what were then current findings in cognitive psychology with the age-old challenges of understanding and designing for developmental opportunities and constraints, Brown raised the question of “what triggers conceptual change” (Brown, 1994, p. 9). Broadly speaking, we think learners must experience a discontinuity between their current state of knowledge and some experiences—direct observations, modeling, manipulation of some representation—that demonstrate the insufficiency of the prior knowledge, along with some kind of dialogue (done internally by the learner as a metacognitive strategy or in interacting with others) that makes public how and why the earlier understanding is either incorrect or incomplete. The implications for designing instruction that facilitates conceptual change, within and across academic disciplines, remains a conundrum for practice, despite generations of changes in curriculum standards. However, Brown cautions, “School practices are influenced by outmoded theories of learning and development that are relics of psychology’s behaviorist past” (Brown, 1994, p. 11). Unfortunately, I would say that is as true today as it was in 1994.
In 1965, Cronbach and, again in 1994, Brown remind us of principles of learning articulated by the Progressive Education Association through the 1940s, which in many ways pushed back against the underlying principles of behaviorism. These principles include the idea of “learning . . . through active practice,” the importance of developmental readiness, the idea that transfer is enhanced with new tasks that are related to prior learning and connected to social practice, the importance of student interests, applying learning to concrete situations, and the importance of “a well-understood verbal generalization . . . favor[ing] the teaching of abstractions, but in close connections with their concrete referents” (Cronbach, 1965, p. 113). But interestingly, Cronbach says, “Each statement or its corollary is partly false, open to dispute, or seriously incomplete, in the light of current research” (Cronbach, 1965, p. 113). In his 1965 address, Cronbach delves into a dilemma of theory to practice that continues even today. With the cognition revolution and its focus on the processes through which people construct new knowledge from observation and experience in the world, what are the implications for problem-based learning in schools, what kinds of scaffolding are required, and what, if any, is the role of rote learning in the context of problem-based learning? This notion of problem-based learning has been particularly central in the design of learning environments in both mathematics and science, with huge investments of federal dollars in small- and large-scale interventions. Findings from such interventions have been mixed, suggesting that Cronbach’s cautions about principles around problem-based learning, learning by doing and discovery, are both true and false. More recent research on learning in and out of school settings, highlighted in Resnick’s (1987) address, offer some useful insights into how contexts and the nature of the problems students are attempting to learn can play in features that are more likely to support learning by doing as opposed to simple rote learning. As Cronbach notes, “I think it pedantic to require children to ‘discover’ that magnets attract iron, but they probably cannot comprehend the properties of a magnetic field save by exploratory investigation” (Cronbach, 1965, p. 116).
Under the behaviorist umbrella, drill and repetition of particular instructional stimulus were aimed at producing a singular response. Under the constructivist umbrella, learning generative concepts and heuristics is deemed an appropriate goal, where instructional practices help learners infer generalizations through direct experience. These continua have also been associated with differential tracking in K–12 schools (Oakes, 1985), where students from communities living in poverty and students from nondominant minorities are more likely to experience instructional practices dominated by rote instruction and where attempts to implement conceptually focused instruction are complicated by the field’s lack of adequate knowledge of how to plan for and navigate the multiple pathways that conceptually focused instruction inevitably demands.
The Shift to Ecological Systems
Snow raises a persistent challenge in her 2001 presidential address: how to understand and interpret displays of competence and to construct generalizations about what such displays mean for policy and practice. She illustrates the argument by drawing parallels between what we know about how young children learn language and what is entailed for new teachers to develop their craft. Learners (whether children or teachers) make errors and construct and test generalizations as they are continuously faced with adapting their incomplete state of knowledge to unexpected challenges. This proposition makes sense as we informally observe young children learning language in everyday contexts. However, in our research on how people learn, the field has a history of categorizing particular errors among particular populations (especially those from nondominant communities) as deficits rather than normative trajectories for learning that entail errors and incomplete knowledge. The field also tends not to raise questions about the generalizations we can validly draw from studies of “errors” students from nondominant populations make in our studies of learning, particularly in school contexts. The field has been less likely to examine such “errors” as incomplete knowledge, inviting us to examine potential links between the formal constructs of interest and the naive conceptions or incomplete understandings that such errors may reveal. In fact, across relevant fields—cognition, applied linguistics, human development, educational assessments, and the broad umbrella of educational studies—researchers have for decades made a well-funded cottage industry of putting forward propositions about cognitive, linguistic, socioemotional, and broadly speaking, cultural deficits that presumably youth from nondominant communities and communities living in poverty bring to school. AERA past president Joyce King (King, Speight, Vaughn, & Vaughn, 2016) calls this phenomenon cultivating “data plantations.” These deficit propositions have been largely directed at participation in formal schooling, in contrast to studies of learning in informal settings. Snow’s characterization of how complex learning unfolds, then, applies not only to how teachers learn their complex practice, as Snow contends; but I propose that Snow’s argument points to a powerful dilemma about human learning to which our science is often not sufficiently sensitive and where the long-standing predilection in our fields to treat diversity in human functioning as arbitrarily hierarchical provides warrants that remain contested.
Snow (2001) goes on to say, “Knowing a language, like knowing how to do research or knowing how to teach, involves going beyond the information modeled by extracting underlying principles and engaging in creative application of those principles” (p. 5). Snow argues this, too, underlies the contingencies inherent in conducting research on learning and, by extension, on practice. I extrapolate from Snow’s challenge the need for research frameworks to take a broader lens to examine the dynamic relations among the multiple factors that constitute the ecological contexts in which humans learn and develop over time. Just as Snow cautions that it is inappropriate to extrapolate foundational principles about effective reading instruction from a single ethnographic study of a classroom, I would argue it is equally incomplete to infer causal generalizations about competence when examining only a restricted range of potential variables, even if one has a national, presumably representative sample of participants, with random assignment, using well-established diagnostic tools for measuring comprehension. This is, in part, because competence is context dependent and influenced by a myriad of other factors—motivation, perceptions of the task and setting, social-emotional states, resources available as sources of support, and vulnerability within the site of testing (e.g. school) as well as within and across the multiple stable sites that make up the ecological niche of the learner(s) (Spencer, 2006).
Thus, I conclude this discussion of the evolving shifts in frameworks for understanding how and what people learn across these seven decades as captured by AERA presidential addresses from each decade with my own presidential address (Lee, 2010). Without question, I was deeply influenced by the shift to a sociocultural framework and the new insights it has provided about the nature of cognition—not purely mental and internal to the individual; distributed across people, artifacts, and the organization of spaces; seeing knowledge and problem solving as diverse in their instantiations within and across cultural communities of practice. However, by 2010 I had also begun to think about the challenges that earlier bodies of research had been articulating since the 1940s around the idea of ecological systems (Barker & Wright, 1949), and in particular, the work and design challenges posed by Bronfenbrenner (Bronfenbrenner & Morris, 1998). In addition, emerging work in human development and the various subfields of the neurosciences—cognitive, social, cultural (Cacioppo, Visser, & Pickett, 2005; Han et al., 2013; Kitayama & Park, 2010; Meltzoff, Kuhl, Movellan, & Sejnowski, 2009; Organization for Economic Cooperation and Development [OECD], 2007; Varma, McCandliss, & Schwartz, 2008; Whitehead, 2010)—along with work on dynamic systems (Fischer & Bidell, 1998; Holland, 1998; Thelen & Smith, 1998; Wilensky & Resnick, 1999) were pushing my thinking. By 2010, I had completed two long-term interventions in schools, based on the framework—Cultural Modeling—I had developed for connecting youth’s prior knowledge from participation in everyday practices, in particular around language use and text genres, to generative heuristics and problem-solving strategies for text comprehension, with a particular focus on literary reasoning (Lee, 1995, 2007). The framework was largely cultural and cognitive in its foundations and proved useful along many dimensions.
However, working on the ground—working alongside practicing teachers in the interventions as a teacher myself—I was confronted with realities that the cultural and cognitive dimensions of the framework did not adequately address: how to understand the multiple demands of adolescent development especially for youth living in intergenerational poverty and minoritized communities who must learn to grapple with both sets of demands; how to understand the sources of resilience on which to draw in design work that went beyond the cognitive demands of tasks; how to understand the multiple influences of school culture, family organization and relationships, peer social networks, and neighborhood resources on youth’s perceptions of their own possibilities and the demands of schooling. I was deeply influenced by Spencer’s PVEST model that conceptualizes navigating sources of risk and resilience as fundamental human tasks, shaped by relations between sources of support and sources of vulnerability (Spencer, Dupree, & Hartmann, 1997). But I was also beginning to investigate research on dynamic systems—both as tools for problem solving and as frameworks for understanding how human physiological and cognitive systems operate in tandem (Fischer & Bidell, 1998). And finally, I was committed to approaching these challenges, particularly with regard to youth living in poverty and youth stigmatized by racism, as fundamental tasks of being human and not as specialized deficits to which a presumed homogeneous middle class or a homogenized conception of Whiteness stands in stark contrast.
Thus in my 2010 talk, I address how biology and culture are intimately intertwined to predispose humans to engage in routine tasks of development (making evaluations of safety and risk, navigating social relationships, solving problems, being resilient, using the tools of language and artifacts from both the natural and the manmade world), how by evolution we are primed to be able to navigate the world through multiple pathways in order to support resilience (Quartz & Sejnowski, 2002), and that people’s participation across multiple contexts matter. And in particular, I sought to raise the importance of attending to physiological process as a source of opportunity and risk, especially for youth living in poverty.
Fundamentally, I argue that to address the complexities and nuances entailed in understanding human learning and to use such understandings to design robust learning opportunities—in school, in families, in workplaces, and in informal learning environments—we need systematic and well-supported opportunities for interdisciplinary collaborations. The world is messy and complex. And certainly we have learned a great deal by stripping away some of the sources of complexity in order to be able to use the methodological and conceptual tools available at any given historic moment, but at the same time, for example, our ongoing dance with how research and theory can inform practice strongly suggests that we need to get a better handle on examining complexity. I have tried to argue in my presidential address some possible pathways and range of domains upon which it might be productive to draw to begin to wrestle with the complexities of learning and development in the real world.
And one final comment with regard to complexity—research in cognition and motivation with additional warrants from across the neurosciences (cognitive, social, cultural) posits that perceptions matter for belief systems, which in turn influence goals. As researchers in education examining how people learn, we must also acknowledge the ecological spaces in which we operate. The long history of educational inequality in opportunity to learn across nations, and in particular within the United States, in tandem with the research traditions that have contributed to these disparities emerges in the context of widely held beliefs about race, ethnicity, and class (as well as other dimensions of difference), or what Charles Mills (1997) calls “the racial contract.” This philosophical and ideological belief system provided what were deemed logical warrants for the eugenics movement (Gould, 1981), which characterized the beginnings of the American Psychological Association. And although in the past few decades it is not deemed politically correct to make explicit attributions of racial deficits, there is still broad-based research that translates difference into deficit, now more likely attributed to poverty; this despite the evidence from high-achieving countries outside of the United States on international measures that poverty in itself is not an inevitable determinant of academic outcomes (OECD, 2010).
Conclusion
The question of how research can inform practice has been a persistent challenge across these six decades and before. I am particularly struck by how, with few exceptions, these presidential addresses focusing on learning do not tackle directly the persistent inequalities in opportunity to learn: whether considering the conditions of schooling under Jim Crow and legal segregation, the persistent achievement gap as measured since 1969 by the National Assessment of Educational Progress as these correlate with race, ethnicity, and class (National Center for Education Statistics, 2013). Certainly, there have been AERA presidential speeches taking on these dilemmas but not in the context of fundamental theorizing about how people learn. In this light, it is understandable that the groundbreaking volume How People Learn, published by the National Research Council in 1991, also did not directly tackle this issue. Common wisdom seems to be that tackling these problems of inequality in opportunity to learn is a question of policy—which certainly it is—but not necessarily a challenge to some of the fundamental propositions we accept about how people learn. We continue to have to wrestle with what are purported to be research-based claims about deficiencies and deficits that learners from low-income communities and from Black and brown communities bring to the tasks of learning in school, despite evidence of plasticity in human development, evidence of resilience in the face of risk and theoretical explanations about sources of resilience, long-term analyses of conceptual links between everyday knowledge derived from experience in the world and formal academic domains, and evidence, for example, of the benefits of bilingualism on cognitive functioning (Carlson & Meltzoff, 2008)—research and public discourse presumed to be informed by research continue the mantra of cultural deficits (Lee, 2009).
AERA has and continues to be a space for continuing to discuss and debate these issues of theory and practice, and of theory informing policy, and for challenging ourselves as researchers.