Some years ago, having toiled my way toward the conclusion of Kavale and Forness's (1995), The Nature of Learning Disabilities: Critical Elements of Diagnosis and Classification, I turned at last to the first page of the final chapter. The title of this final chapter read — What is a Learning Disability? Initially, I was bemused, and then amused, by the thought that one might reasonably have expected the preceding several hundred pages to have answered that all important question. Nevertheless, having gone thus far, I was committed to hearing them out. As I read on, I was instructed that, "LD is a real phenomenon," and warned that, "It is counterproductive to continue asking whether LD is real" (p. 333). Amusement turned to incredulity when I arrived at the following complaint:

It should be understood that objections to the LD category stem from our own inability to understand LD. The ceaseless empirical, political, and philosophical debates in the LD field have undermined the LD concept, leaving it vulnerable to questions about its existence. Instead of discussing ways in which the LD concept might be developed in a logical and rational fashion, we have been reduced to bickering over whether the object of our discussion even exists. After some 25 years, this is not the level of debate that fosters confidence that we know what we are talking about [emphasis added]. (p. 333)

The origin of my incredulity was not the authors' failure to answer the question that formed the raison d etre of their book — "What is a learning disability?" I had long ago concluded that the "condition" in question was not a "thing" people have, but instead an interpretation of difference in a particular cultural context (see: Carrier, 1983, 1986, Coles, 1987, 1989; Klatt, 1991; Sleeter, 1986, 1995, 1998). Rather, I was stunned by the authors' (presumably inadvertent) blunt admission that they could not, after all, supply an answer.

I say presumably inadvertent because it seemed to me that this admission should have inspired considerable pause and reflection on their parts. My initial assessment was somewhat off the mark, however, because a great deal of reflection, of sorts, has indeed taken place among special education academics both before and since the inclusion of the learning disabilities category in the Education for All Handicapped Children Act (PL 94-142; 1975). The problem is that their inquiry has been regimented and delimited by their understanding of the meaning of theory in the social sciences. More to the point, their theoretical commitment to understanding learning disability as a real condition intrinsic to the individual; and, by extension, presumably neurological in origin has led to years of bewildering deliberations about how to resolve the contradictions and stumbling blocks created by these commitments.

Carrier (1986) characterized this theoretical and ontological commitment quite clearly, stating that:

Learning disability theory presents itself as a set of scientific statements about reality, as an assertion that the world is a certain way and not some other way. The growth of the theory occasioned and was occasioned by the spread of the idea that learning disability is a real condition with real causes and real effects; that certain attributes of certain children are the signs or symptoms of a learning disability which springs from a disordering of the child's basic thought processes, brought about by a neurological malfunction of some sort. And advocates of the theory were quite aware of the need to establish the construction of reality embedded in learning disability theory. (p. 55)

That learning disability theorists have been politically successful in establishing learning disability as a real condition is made evident by the fact that, over the years, millions of school children have been identified as having this condition.1 As well, most educators and the vast majority of the public are convinced that people with learning disabilities possess brain defects that cause them to misperceive visual and auditory information, reverse words and letters while reading, and so on.2 Success notwithstanding, maintaining learning disability as an established condition has required ongoing attempts at reconstruction. The field's ongoing tribulations vis-à-vis efforts to define and redefine learning disabilities while simultaneously maintaining their objective/independent existence has constituted something of a quiet crisis in that these activities should long ago have signaled some rather momentous implications raising fundamental doubts and well-placed objections from educators and the public alike.

It is no small irony that this situation has been hiding in plain sight for decades. While children are routinely identified and labeled as having learning disabilities, efforts have been ongoing to reconfigure and reconceptualize the criteria and hence the very nature of this disability. In what might well be seen as an act of some desperation, learning disability's most recent incarnation centers on those students who fail to respond to evidence-based practices. This Response to Intervention (RTI) model (see: Bender, 2007; Kauffman & Hallahan, 2005) represents yet another effort in the chase to outrun the flaws in earlier learning disability identification practices, thus preserving a particular way of seeing the world by repackaging its "theoretical" perspective to keep it alive.

In what follows, I outline the development of learning disability theory through an analysis of efforts to inaugurate viable criteria for identification, attending particularly to internal controversies among leading learning disability scholars surrounding the new RTI model. Next, I use this analysis as a vehicle to discuss the larger problem with theory in the social sciences and conclude by exploring alternatives to theory as commonly understood.

The Emergence and Development of Learning Disability Theory

In the beginning, there were children who exhibited what was then (and now) referred to as "unexplained underachievement" (Fletcher, Morris, & Lyon, 2003 Mastropieri & Scruggs, 2005; Johnson & Myklebust, 1967; Strauss & Lehtinen, 1947). Quite simply, these were (and are) children who seemed bright enough but nevertheless could not keep up with their classmates academically. An explanation for this enigma was in order. It was subsequently delivered in the late 1940s by neuropsychiatrist Alfred Strauss who inferred that such children must be brain damaged. He drew this inference because their behavior on the perceptual tests to which he submitted them matched the behavior of individuals known to have sustained a brain injury. Accordingly, Strauss reasoned that because people known to have experienced an actual injury demonstrated certain behavioral characteristics, anyone exhibiting those characteristics must themselves be brain damaged (see: Carrier, 1986 for a detailed historical/analytical account). When critics soon pointed out the fact that this diagnosis involved an erroneous leap of inferential logic (e.g., Sarason, 1949), Strauss ignored their concerns and reasserted his supposition on the grounds that, faulty logic notwithstanding, it was one of promising utility (again, see: Carrier, 1986). In plainer terms, it was a satisfying explanation for him, though one could hardly call it scientific. As we shall see, Strauss's evasion of legitimate critique set a pattern that has continued to this day.

By the early 1970s the neurological damage framework revealed a number of disconcerting fault lines. Paramount among these was the thorny matter of sustaining the assertion of brain damage when critics pressed the fact that no actual physical proof of such existed (Benton, 1973; Birch, 1964). This task became all the more difficult when tests revealed that "normal" children exhibited the very same "soft signs" of neurological damage [at this point referred to as minimal brain dysfunction] as those experiencing unexplained underachievement (Adams, Kocsis, & Estes, 1974; Copple & Isom, 1968; Hart, Rennick, Klinge, & Schwartz, 1974). Conversely, some children with unexplained underachievement showed no "soft signs" of neurological damage when tested (Benton, 1973, Birch, 1964). One explanation for these confounding revelations was attributed to yet another hitch in the project — the tests used to detect faulty neurology were deemed lacking the necessary technical adequacy (Herbert, 1964; Yates, 1954), although one might reasonably inquire how the "validity" of such tests could in fact be established.

Subsequently, a similar set of impediments had begun to emerge on a more practical front. What if unexplained underachievement could be attributed to psychosocial factors instead of neurobiological ones (Forness, 1982; Mayron, 1978)? Some youngsters, for example, arrived at school too upset by their life circumstances to benefit from teaching. Although the most widely accepted definitions of learning disability specifically excluded children whose learning problems were deemed to be caused by cultural, environmental, or economic disadvantage, these "exclusion clauses" failed to cover a multitude of very good reasons for why apparently "bright" children fell short of their "potential." Poor instruction could also account for underachievement (a point of contention related to the new RTI approach discussed in a latter portion of this paper). Practically speaking as well, it became obvious that students identified as having learning disabilities differed more from each other than from students not identified as such (MacMillan, 1973). Thus, what would be the point of identifying and segregating these students for remedial instruction if they did not constitute some form of homogeneity? Finally, and even more troublesome from a practical perspective, attributing learning disability to presumed neurological damage or minimal brain dysfunction offered educators little or no guidance to inform teaching practices (Bateman, 1974).

Issues raised in the 1980s and 90s served only to compound the murkiness of learning disability theory. The swelling multitudes of identified students inspired a preoccupation with improving the learning disability definition and criteria for identification that endures to this day (see for example: Bradley, Danielson, & Doolittle, 2005). This tinkering-with-the-mechanism-mentality preserved the ontological presumption of learning disability's objective existence, thus staving off any meaningful scrutiny of its theoretical underpinnings. Surely there were some rock bottom criteria for identifying students who really did have a learning disability, thus distinguishing them from those who did not. Once these enigmatic criteria were discovered, the true definition of learning disability would rise up out of the shadows and reveal itself at last.

The acknowledgment that learning disability could not explicitly be distinguished from mild mental disability and behavioral disorders has constituted a further and telling source of conceptual instability. Though academic underachievement serves as a common hallmark of all three categories, the putative underlying causes of underachievement specific to each category were meant to distinguish one disability from the other. Simply put, those with mental disabilities are not intelligent enough to keep up with their "normal" peers while those with behavioral disorders, though intelligent enough, cannot benefit from instruction because they are too distracted by their emotional and behavioral problems. When the criteria for identifying mental disability and behavior disorders became more stringent throughout the 70s and 80s, the learning disability category was left holding the bag, so to speak. Underachieving students who no longer met the stricter criteria for mental disability or behavior disorder spilled into learning disability programs, swelling the ranks of this category in disconcerting numbers.

Kavale and Forness (1998) blamed the "demonization" of IQ tests as culturally and racially biased for raising the threshold for identifying students with mental disability. For them, unwarranted political and judicial influence resulted in fewer students being identified as having mental disabilities, the consequence of which was the muddying of the learning disability waters by improperly placing the "garden-variety" underachievers in special education programs as students with learning disabilities instead. Similarly, they pointed to the growing political reluctance to pin students with the stigmatizing BD label as resulting in stricter criteria for definition of this disability category. Consequently, students who previously would have been identified as BD wound up being identified as having a learning disability. Lamenting learning disability's resulting "loss of integrity," they described the category as, "a sponge wiping up the spills of general education" (Kavale & Forness, 1998, p. 250)

On another front, Kauffman, Hallahan, & Lloyd (1998) attempted to defend the integrity of learning disability by impugning the viability of mental retardation. Their argument went as follows:

Although learning disabilities may be unfavorably compared with mental retardation on the dimension of diagnostic clarity (under the assumption that the latter has a clear, formal demarcation from normalcy that is lacking in the former), it is instructive to note that the discrimination between mental retardation and normalcy is not a fact of nature but one that we set arbitrarily. (279)

Apparently failing to comprehend the vulnerability of their position, their implicit defense of the learning disability category appears to be their presumption that the category is indeed a "fact of nature" and therefore not one "we set arbitrarily." Thus learning disability could be rescued in the act of throwing mental disability under the proverbial bus.

Throughout, the strongest case for the "existence" of learning disability remained centered on the discrepancy between students' performance on standardized measures of achievement and intelligence. Such a discrepancy, it was assumed, occurs because the student experiences neurologically-based difficulties in processing information. Various statistical formulas widely used to detect a "significant" discrepancy soon fell upon hard times, however, due to statistical, computational, and other technical inadequacies, some of which continue to defy renovation (Reynolds, 1985). But these technical inadequacies were far from the most intractable aspect of identifying learning disability using the discrepancy model. In his extensive review of the research on IQ- achievement discrepancy, Aaron (1997) concluded:

The studies reviewed so far overwhelmingly undermine the validity of the premise that there are qualitative differences between LD and non-LD poor readers in cognitive processes that are intrinsic to reading. Even reversal errors in reading and writing, once considered a marker for developmental dyslexia, are reportedly equally prevalent among LD and non-LD poor readers….The very precepts on which the edifice of LD has been constructed here stand challenged. (p. 470, emphasis added)

One might further suggest a certain incongruity in pursuing the discrepancy criterion in the event that an individual's neurological processing deficits affected not only his performance on the achievement test but also on the IQ test. Nonetheless, Aaron's excoriating analysis of the discrepancy model left no reasonable alternative to its abandonment. His indictment pointed beyond technical difficulties to the original flaw in learning disability theory.

For the duration of these trying times, the field of learning disability managed to ward off serious threats to its survival. Its Teflon surface continued to deflect what should otherwise have been annihilating critique by critical theorists (Carrier, 1983, 1986; Sleeter, 1986, 1995, Tomlinson, 1988) and detractors both within the field of special education (Algozzine & Ysseldyke, 1983; Poplin, 1984, 1987; Skrtic, 1999) and without (Coles, 1987; 1989; Klatt, 1991). Learning disability theorists' routine confessions that they themselves did not know what a learning disability was did not preclude equally routine condemnation of critics who committed the unthinkable transgression of making too much of the problem. For example, in their rebuttal of Sleeter's (1986) critique of learning disability, Kavale and Forness (1987) declared that her bent for revisionist historical interpretation was nothing short of fallacious, adding without a hint of the preposterous that "perhaps such controversy might be avoided if we could answer more scientifically the question 'What is a learning disability?'" (p.11). One of the critics captured the dazzling element of farce in the situation by stating that

The field of "learning disability" flew off like Hippogriffe into the heights of scientific respectability and has been subjected to unrelentingly severe criticism, but to this day continues to march on vigorously like Saint Denis with his own head carried in his hands. (Klatt, 1991, p. 59)

Strident insistence aside, preservation of the learning disability category might well have been most effectively facilitated by theorists' success in sequestering critique and setting it on their own terms, thus creating the impression that the problem was imminently resolvable given adequate time and attention. It is to their most recent attempt to repackage learning disability that I now turn.

Response to Intervention: Learning Disability for the Twenty-First Century

As the turn of the 21st century approached, stratagem for deflecting critique had begun to wear thin. Failed efforts to establish a coherent theory of learning disability fundamentally threatened the extinction of the field. Still, the prospect of "throwing the baby out with the bathwater" (see: Scruggs & Mastropieri, 2002) struck some as anathema. To do so would require dismissing out of hand a long history of "descriptive reports dating back to the 19th century or earlier…the proliferation of professional and advocacy organizations….and the thousands of service providers who not only testify to the existence of learning disabilities but also advocate for improved services" (Scruggs & Mastropieri, 2002, p. 158). Transforming a social reality into an "objective" reality is hard work, but somebody had to do it.

Learning disability theorists proposed a surfeit of alternative criteria and models for pinning down the true nature of the condition. These alternatives included: Double-deficit Criteria, the Phonological Process Core Difference; Chronological Age Definitions, Bayesian Procedures, Neuropsychological Assessment, Assessment of Cognitive Processing, Operational Interpretation, and, finally, Failure to Respond to Validated Treatment Protocols (Dual Discrepancy Criteria) (see Scruggs & Mastropieri, 2002 for a full description of each). Although time and space does not permit a thoroughgoing examination of each of these alternatives, a discerning appraisal of their descriptive titles reveals their fated kinship to identification models and criteria previously put to the test and found wanting. Providing even more optimism for ultimate vindication was the ongoing suggestion that while research had not as yet confirmed neurological and/or genetic causes of learning disability, such a possibility could not in any event be ruled out (Hynd, Clinton, & Hiemenz, 1999; Olson, 1999; Rumsey, 1996; Wood & Grigorenko, 2001). Nevertheless, a version of the last approach listed above (failure to respond to validated treatment protocols) appeared dynamically novel enough to warrant serious investment, at least until the long sought after proof of neurological pathology at last emerged.

The apparent logic behind the new Response to Intervention (RTI) model is reminiscent of the original learning disability theory in that it relies on inferring the presence of learning disability based on interpretation of behavior. Those who do not respond to instructional interventions "scientifically" proven to be effective must accordingly have the disability. At present there are various approaches to implementing RTI and, although its use has already been written into U.S. federal legislation, there appears to be significant disagreement not only on how it should be put into practice, but more fundamentally whether it should be used at all.

Some leaders in the field of learning disabilities recommended moving "away from locating problems within the learner and toward a process that focuses on improving student achievement by improving the overall instructional process" (Bradley, Danielson, & Doolittle, 2005, p. 486), while others were intent on retaining learning disability as an intrinsic impairment (Gerber, 2005; Mastropieri & Scruggs, 2005). The latter warned that the RTI model, unless accompanied by an IQ test, fell short of putting a lock on the impaired neurology inference, even if the insinuation of such impairment remained. And while virtually all admitted that a huge range of questions about the feasibility of RTI remained unanswered even as the model had already been implemented, some were clearly more concerned about those unanswered questions than others.

Among the more concerned, Gerber (2005) took an "economics" approach to the issue, pointing out that teacher "tolerances" and resources are ultimately the meaningful test of which students will be identified as learning disabled. More directly, RTI would not deter intolerant, under-resourced, and/or over-worked teachers from discovering learning disabilities wherever they may lurk. He further warned that not only is RTI prohibitively expensive to use on a large scale, it is also naïve to believe that evidence-based practices will be as rigorously applied in the everyday classrooms as they are in experimental settings. Finally, he cautioned that a move toward RTI would detract from what he believed was growing evidence that there are in fact some students whose learning problems "are associated with a clear neurological substrate" (p. 522). And, while allowing that RTI could be useful in identifying and helping students who needed help, he added that, "this approach will not, by itself, put us closer to understanding learning disabilities" (p. 522, emphasis in original).

Kavale (2005) also counseled prudence a propos leaping headlong into RTI as the answer to the learning disability identification conundrum. Pointing out that the current RTI interventions were too narrowly focused on phonological processing and currently enjoyed only modest "empirical validation," he intimated that making RTI the centerpiece of learning disability identification was more than a bit precipitous. Rather than stepping into that great unknown, he proposed a back-to-the-future approach, arguing that,

With modern theories about the importance of processing skills replacing outdated views (e. g., perceptual motor deficits) that were associated with the SLD [specific learning disabilities] concept when it was first proposed, new validated process deficits should be incorporated in an operational definition of SLD. (p. 554, emphasis added)

Of course, this seemingly forward-looking solution requires a certain amnesia concerning Strauss's original tautology. Invoking new theories about processing skills manifestly requires the same specious logic as the earlier neurological impairment inference. Likewise, Kavale encouraged a resurrection of the discrepancy model, with the proviso that it simply be more rigorously and consistently applied. RTI supporters, he claimed, "appear to have exaggerated the deficiencies presumed to be associated with discrepancy" (p. 554). Even longtime critics of the learning disability category can admire the shrewdness involved in his insistence on new and improved versions of traditional identification criteria. Such a strategy allows conservationists to buy time (perhaps indefinitely) while the burden to expose the ineluctable cracks in the foundations of the new and improved is neatly shifted to others.

Given the substantial and assiduously mounting literature on RTI, one could extend the above commentary at vertiginous length and with corresponding tedium. Doing so would be pointlessly redundant in making the case that learning disability scholars' implacability is matched only by their by the depth of theoretical confusion. As Thomas (2007) perceptively observed, "Theory, in a broad variety of forms…seems somehow to have the power to curtain reality and the likely practical consequences of 'implementation' from the eyes of academics and practitioners who otherwise seem quite sensible" (p. 10). Just so, which prompts us to question — what is it about the way theory is understood and plays itself out in the social sciences that leads to these kinds of quagmires?

The Problem with Theory in the Social Sciences

Carr (2006) rightly situated the rise of educational theory as an artifact of the Enlightenment, adding that "it is only in the light of the epistemological assumptions of modernity that we can adequately understand the key episodes in the twentieth-century educational theory debate" (p. 143). Enticed by the conspicuous successes of physical scientists' mastery over the natural world, social "scientists" could scarcely withstand the alluring possibility that such intellectual and practical achievements could also be theirs (Giddens, 1976). A vital component of this mastery was the centrality of theory-building. Thus the social borrowed from the physical both its methods and its methodology; and, as part of the latter, it bought into a particular account of theory, the central elements of which required that scientific theories lend themselves to empirical tests and refutations, produce objective knowledge, and lead to law-like generalizations which warrant proper counterfactual conditionals for the ultimate purpose of prediction and control. The theorist, in other words,

by virtue of a withdrawal or standing-apart from the spectacle, can see the whole of the spectacle as a whole and without the constraints of partiality and particularity that are the definitive conditions of enacting a part in/of the spectacle. By withdrawing from any particular part and interest, [he or she] achieves a 'death' of the contingent self, leaving in rebirth a kenotic presence that is free to behold in abandonment that which is unconcealed. (Hazelrigg, 1989a, p. 14)

That the sociological (or educational) theorist is unable to achieve this special way of seeing, is unable to accomplish the "death of the contingent self," has been revealed not only by those like Popper and Merton who, their critiques of theory notwithstanding, wished to preserve the scientific status of social sciences but also, and far more devastatingly, by those like Gadamer (1975), Rorty (1979, 1989, 1991), and Hazelrigg (1989a, 1989b) who envisioned the social sciences altogether differently.

In his well-known critique of positivist verificationism, Popper (1959; 1963) acknowledged that human observation is always selective and therefore values-laden. Accordingly, this insight meant that theories could not be empirically verified. To avoid the consequences of this situation, he attempted to rescue his version of scientific realism by replacing the requirement of verification with that of falsification, i.e., a theory could be considered scientific until it could be falsified or at least in principle be falsified. By this standard, he viewed Freud's psychoanalytic theories to be pseudo-scientific because they could not, even in principle, be falsified (see also, Bernstein, 1976). Simply put, he saw that these theories were immune to falsification because their flexibility allowed them to explain any form of human behavior. In explaining everything, they explained nothing.

Like Popper, Merton (1949) saw a number of flaws in sociological theory, but also sought to recoup the possibility that, like the physical sciences, the social sciences could engage in theory-building. They must, it seemed, be able to claim theory-building. Otherwise the social sciences could not be cumulative and thus would be forced to abandon the authority and cache of science. Well aware of the deficiencies of sociological theories writ large, Merton turned a critical eye toward what he termed "post factum sociological interpretations." Bernstein (1976) described these pseudo theories as "a variety of crude hypotheses which are in some measure confirmed or verified by the 'facts,' but which are designed to account for conflicting and contradictory states of affairs," resulting in theories which are "frequently so flexible, vague, or open that they can 'account' for almost any data" (p. 9-10).

It does not require deep analysis to see the connection between Popper and Merton's critiques and long troubled history of learning disability theory. Clearly, because learning disability theory, like Freud's psychoanalytic theories, cannot be falsified and because it qualifies as an unambiguous example of what Merton described as a post-factum interpretation, learning disability theory does not hold up even by the lights of those who hoped to preserve theory's scientific status. But its problems do not end there. Its fate is inextricably tied to that of all sociological/educational theory in the wake of further philosophical scrutiny.

Hazelrigg (1989a), for example, deftly closed off Popper's falsification thesis simply by demonstrating the full implications of what Popper himself had already accepted — all observation is values-laden or theory-impregnated. "First of all," he queried, "if observations and facts are theory-impregnated, why should any given observation or fact, or any set thereof, be accorded the significant virtue of being able to falsify a theory?" (p. 78) There must, in other words, be some theory-free fact equal to the task of explicitly "falsifying" the theory in question. Otherwise, how might one know that the theory had indeed been falsified? Was the falsification in question a true falsification or a false falsification?

In like manner, Bernstein (1976) confirmed that the viability of all foundationalist sociological theories (including Merton's more modest proposal of theories of the middle range) hinges on the possibility of distinguishing facts from values, a distinction rendered irretrievably untenable by the impossibility of theory-free observation (see: Hanson, 1958; Nagel, 1986; Putnam, 1981; Rorty, 1979). Consequently, social theories inescapably wind up presupposing the facts they purport to explain. In the case of learning disability, we encounter the absurdity of presuming its status as an ontological fact in the course of theorizing about what "it" is. Accordingly, Bernstein (1976) concluded that there is no bona fide pattern of conjecture and refutation to be found in the social sciences. Instead, there are only general orientations that have their time in the sun only to be supplanted by newer and seemingly more promising ones. But because it "bears a superficial resemblance to the physical sciences" (p. 26), this pattern moves forward impervious to the telling criticisms that are now so common in the philosophy of social sciences literature.

Under these circumstances, Carr's (2006) conclusion that "educational theory has run its course and should now be brought to a dignified end" (p. 136), is right on the mark if the foundationalist version of theory he so capably undermines is all that is left to us. For what he pointedly showed us is that this understanding of theory is a "flawed intellectual project" (p. 137) directly responsible for the social sciences' now longstanding failure of achievement. What then is the purpose of the social sciences without theory as commonly understood now that it can no longer serve as the standard bearer of what social inquiry is all about?

An Alternative Understanding or Who Cares About Theory?

We all see the world from a historically and culturally contingent place in it — through particular lens — hence the concept that there is no theory-free observation and thus no theory-free knowledge. Researchers seek explanations and meaning in an attempt to makes sense of things, to anticipate the results of their actions, and generally to pursue the question — how do things work? And because they view the world from a particular place in it, they are disposed to see some things, not see other things, or to "theorize" about things in certain ways. They cannot carry on otherwise because inquiry both shapes and is shaped by researchers' historical contexts, intentions, and predispositions. Researchers' observations are undeniably framed some ways rather than others, whether we call their explanations about how things work theories, explanatory frameworks, or some other term. I agree with Thomas (1997) that the word theory is at this point so hopelessly encumbered by conceptual baggage as to be meaningless, but to get too caught up in semantics may be, as some might say, to scratch where it doesn't itch. The concept of theory as imported from the physical sciences does not work for us — the tinkering of Popper and Merton notwithstanding.

The search for foundationalist knowledge and, by extension, foundationalist theory was and is the seeing of the world in a particular way. It was and is framed by researchers' attempts to claim value neutrality, to achieve objectivity, and to position themselves, above all else, as non-ideological. The story they tell about themselves depicts them as the neutral brokers, the social engineers whose work makes constant progress toward achieving social ideals and goals. Yet, "despite all the talk of objectivity and value neutrality, social science literature and so-called empirical theory are shot through with explicit and implicit value judgments, and controversial normative and ideological claims" (Bernstein, 1976, p. 52-53).

To put it in Gadamer's (1975) terms, their depiction of themselves as non-ideological is in itself a form of ideology — a prejudice against prejudice. And this particular prejudice has consequences in that,

A person who believes he is free of prejudices, relying on the objectivity of his procedures and denying that he is himself conditioned by historical circumstances, experiences the power of the prejudices that unconsciously dominate him…a person who does not admit that he is dominated by prejudices will fail to see what manifests itself by their light." (Gadamer, 1975, p. 360, emphasis added)

In making a distinction between what he called enabling and disabling prejudices, Gadamer repudiated the notion of prejudice as human deficiency, turning instead toward the understanding that prejudices (understood as pre-judgments) are on the contrary what make it possible to make sense of the world once we drop the notion that we can be non-ideological or unprejudiced.

From this vantage point, the pursuit of knowledge or theory (if one wants to use the term) is not about cordoning off prejudice. Because such a cordoning off cannot be accomplished in any case, believing that one can or should attempt to do away with prejudice becomes an impediment. In a far more refined and detailed manner than can be readily summarized here, Gadamer calls on us to be open to each others' questions and doubts, to be aware that we are historically situated and limited in our understandings, and to be conscious of our finite ability to know. Only then, by "risking our prejudices," can we ask the right questions so that our prejudices can become enabling ones rather than disabling ones. This reappraisal of what it means to inquire, or to "theorize," reframes these activities as moral and political endeavors.

In like manner, Rorty (1999) urged us to surrender the thoroughly misguided notion that the purpose of inquiry is to achieve accurate depictions of "reality" so as to distinguish social constructs from other "things." For him, the point of this surrender was to get on with the crucial responsibility of "debating the [moral] utility [consequences] of alternative constructs" (p. 86). In plain terms, the recognition that human constructs, explanations, descriptions, theories, or whatever we choose to call them inescapably sub-serve moral interests and purposes (see also, Putnam, 1981) requires an accounting of the consequences of holding one construct over another.

Theory in social inquiry, if one still wants to use the term, should be seen for what it is and has always been — a moral stance, a preferred way of seeing the world. The guiding metaphor is not one of discovery, but of constructing (Hazelrigg, 1989a, 1989b; Rorty, 1989, 1991). To conduct research or inquiry is to engage in a knowledge-making activity. Researchers construct the social and educational world; they do not find or discover it. In shifting from the metaphor of discovery to one of constructing, researchers acknowledge that their ways of describing things (or people) are inevitably bound to their intentions, goals, and values. One is morally responsible for what one constructs. The question then becomes: what are the consequences of holding or believing one construct over another? What is gained or lost in "theorizing" one way as opposed to another? Or, per the present discussion about learning disability, who wins and who loses when students' "unexpected" underachievement is described as neurological deficiency or intrinsic impairment?

Theories about "Learning Disability" or What is Better for Us To Believe?

The field of special education as a distinct professional enterprise is fundamentally dependent upon disability theorized as an objective condition intrinsically owned by (and manifested within) individual students. What this patently means is that special education owes its status and legitimacy to its success in accomplishing what Danziger (1990) referred to as "constructing the [psychological] subject," or more to the point, the creation of conditions (educational disabilities) by which professionals could promote their methods and interventions (Brantlinger, 2004). These methods and interventions have been favorably received and supported because:

The characterization of individuals in terms of their degree of conformity to criteria of performance within specific social contexts made it easy to redefine the problems of institutions, or even the problems of society, as individual problems. If all social problems were nothing more than the aggregate of individual problems, they could be handled by appropriate treatment of individuals and required no questioning of the social order. (Danziger, 1990, p. 109-110)

Though none of this impugns what can be understood as a sincere desire to help students struggling in school, it does suggest, and suggest strongly, that the desire to help can also be imbued however unconsciously with decidedly less worthy aims and goals (Thomas & Loxley, 2005).

It has been some time since critical theorists such as Carrier (1986) and Sleeter (1986) drove a moral stake through the heart of the medicalized version of learning disability. Resonating with Danziger's observation that "constructing the subject" precludes what to some is untoward interrogation of the existing social order, both made the case that the creation of learning disability was nothing short of a sanctioning of social and educational inequality. It has also been some time since other critics took aim at the scientific pretenses (Algozzine & Ysseldyke, 1983; Coles, 1987; Klatt, 1991) and epistemological frailties (Poplin, 1984; Skritc, 1999) of learning disability theory. On the whole, all of these critics in one form or another summoned educational researchers to reexamine learning disability within a context of integrity, justice, and fairness.

More recently, Kim Reid and Jan Valle (2004) analyzed learning disability as a product of discursive practices "that both define people as having LD and determine what happens to them after they are so labeled" (p. 466). In skillfully making the case that learning disabilities are, and can be nothing other than, a cultural construction, they concluded that it is nothing short of an ethical obligation to change the context of schools such that students' learning differences need not be interpreted as disability. Using a compelling example of classroom dialog, Curt Dudley-Marling (2004) likewise provided a close angle lens on how learning disability can be constructed and un-constructed depending upon teaching practices informed by teachers' predispositions to view students as either incompetent or competent learners. If pedagogical practices play a leading role in creating and un-creating learning disability, the condition is not something that just happens, something that has its own ontological existence. Contemporary intersectionality scholarship has further confirmed the interlocking and structuralized forces of racism, ableism, and classism at work in the overrepresentation of poor and minority students in special education (see for example: Connor, 2008; Ferri, 2004; Ferri & Connor, 2006). This scholarship has demonstrated the synergistically confounding effects of disabling prejudice run amok when the porous boundaries of socially constructed categories borne of bigotry and intolerance morph into one another.

So, what is better for us to believe about learning disability? How do we "theorize" unexpected underachievement? One choice is to soldier on in an effort to pin down with finality the elusive defect within, to persist in believing that research and its theories discover objective facts and realities. Doing so invites more of the same discrediting incoherence that leaves researchers and educators in a state of continual befuddlement. It leads to more harm to individuals who in the tender years of childhood are saddled with demeaning labels and segregated from their peers. In short, it perpetuates rather than ameliorates educational and social inequality.

Or, we can understand learning disability as a self-immolating description, a destructive narrative that has preempted other, more edifying, discourses (Rorty, 1979). This alternative leads us to ask more useful questions and arrive at better, more satisfying and morally defensible answers. Thus, the conversation can then be turned toward understandings of human difference that lead to more productive teaching practices and arrangements, more inclusion, and, in the end, a more lucid sense of what we are about as educators and researchers.