Educational research is hard science...very hard

In a 2006 TED talk, Ken Robinson called on us to “rethink the fundamental principles on which we’re educating our children.” Specifically, he wants to change how we think about creativity. Robinson passionately believes that “we don’t grow into creativity, we grow out of it.” If so, then the problem for educators like me is not so much to explicitly teach creativity but instead to avoid killing it. Robinson supported this claim with a quote from Picasso who seems to have shared Robinson’s passionate belief. Ken Robinson is charming and very funny, and he attacks a movement unpopular among educators to focus schooling on a core of subjects at the expense of the arts, so I am very much inclined to want to agree with him, but I nevertheless wonder if what he is saying true?

If creativity is, as Robinson later defined it, “the process of having original ideas that have value,” which come about “through the interaction of different disciplinary ways of seeing things,” then education seems to me to be essential to creativity, and schools must do more than get out of the way of inherent human creative faculties in order to foster it. At any rate, in order to adjudicate between competing passionate beliefs on education for creativity, what we need is data rather than anecdotes and quotes. I entered this program because I passionately believe that the fundamental principles upon which we educate ought to be informed by research rather than on the passionately held beliefs of the most charming speakers with the most moving anecdotes or based on consistency with political ideologies

When I first applied for teaching jobs 17 years ago, I was frequently asked to include a statement of my philosophy of education. It was to be a statement of what I passionately believe concerning how students learn best. Today the idea of asking for such a statement sounds quite strange to me since researchers have accumulated a lot of actual knowledge about how students learn best. We can approach claims scientifically to help distinguish true beliefs from cherished ones no matter how passionately believed, and we ought to try to do so. That is what I hope to do as an educational researcher.

Waiting for Lex Luthor Part 3

The most commonly cited test score evidence in support of today’s doomsday predictions come from the Trends in International Math and Science Study (TIMSS), a study that tests students throughout the world every four years and compares education systems by nation. U.S. News reported on the latest results in 2008 that “the United States has failed to raise student achievement in science over the past decade while Singapore and several other Asian countries continue to score higher in both subjects.” You may have heard such well-publicized claims that the U.S. is lagging behind Singapore, South Korea, Taiwan, Hong Kong, and Japan in math and science achievement as measured by these tests. U.S. News concluded that “the results suggest that the United States is not doing enough to train the next generation of scientists.” They do indeed seem to suggest that, but that claim is probably false

A researcher named Erling Boe at the University of Pennsylvania made an interesting finding using the TIMSS data. In addition to mathematics and science questions, these tests also ask more than one hundred biographical questions about the students and their families. As you might imagine, students asked to take the test frequently can not be bothered to answer some or even any of these questions, and the average number of biographical questions varies from nation to nation. American students, it turns out, tend to be less likely to feel obliged to answer these questions than students in other industrialized nations. It also turns out that the countries where students are most likely to be diligent in answering these preliminary questions score the best on these same tests.

When Boe ranked the nations in terms of the average number of biographical questions answered, he found that the rankings were the same as those for the TIMSS math test results. In other words, it is possible to rank the nations in math test performance by gauging their persistence in answering boring questions about the number of siblings the students have, their parents’ occupations and education, etc. and only based on such questions. If these same rankings can be formed without asking a single math or science question, we should be quite skeptical about believing that the TIMSS rankings are comparing math and science achievement at all. 

Boe’s study illustrates that what the test may be telling us is that students in Singapore, South Korea, Taiwan, Hong Kong, and Japan are more willing to work hard on a test that has absolutely no direct consequences for them whatsoever just because an adult told them to try their best, and that is perhaps all the TIMSS rankings say.

If it is the case that differences we see between nations can largely be explained by willingness to comply with adult requests for no other reason than that they asked for it, then is an inferior ranking a good thing or a bad thing? I think there are pros and cons to weigh. It is generally a good thing for young people to be obedient to their teachers, but perhaps like me you can easily imagine Bill Gates, Steve Jobs, and Mark Zuckerberg as having once been the sort of high school students who would not have seen fit to give their best to such a test. Perhaps the rebellious tendency of American youth as compared to other nations is part of what contributes to the fact that we produce so many great innovators.

Waiting for Lex Luthor Part 2

We may all very much like to know the answer to George Bush’s infamous question, “Is our children learning?” but the answer will not come as easily as a comparison of standardized test scores across schools, states, national borders, or time. To understand some of the difficulties in comparing different countries to decide which ones are on the way to displacing the U.S. as an economic super power, consider how hard it is just to assess progress in our own country. 

The “A Nation at Risk” report based its alarmist claims about our failing schools on declining SAT scores, but they failed to take into account the fact that the population of SAT takers has changed significantly over time. Because of social changes in the sixties and seventies, by 1983, it was no longer the case as it had been in the past that only a relatively small proportion of students pursuing future education in elite schools were taking the test. A natural consequence of increasing the proportion of students who go on to post-secondary education (an obvious real good) is lower average SAT scores (a merely apparent bad).

By the same token, ranking educational systems by comparison of state’s average SAT scores is just as problematic. Pennsylvania ranked 42nd in the nation for average SAT scores in 2010 with its average score of 1473. Arkansas, which currently ranks 14th, has a far higher average score of 1684. This more than 200 point difference can be explained by the fact that 71% of students take the SAT in Pennsylvania while only 3% of teens in Arkansas do.  Controlling for the effect of the percent taking the test, a simple linear regression reveals that PA is actually scoring better than Arkansas on the SAT (and not surprisingly, Pennsylvanian’s outscore Arkansians on the less popular ACT, the test that 81% of Arkansians take), but let’s still not jump to any conclusions. Consider that the percent taking the test is just one lurking variable which was not taken into account in the original naive comparison that we happen to have been clever enough to identify. Who knows how many other factors may also be important? There is just no way to know how many other such variables ought to be controlled to avoid making naive judgments about states’ education programs based on SAT scores. We know it is easy to be naive, but we can have no idea how hard it is to be wise.

We have similar difficulties in comparing schools to other schools. Consider that there is a very simple way that a high school can improve its school profile with an increased average SAT score. Instead of seeking ways to increase the proportion of its students who pursue higher education as they generally do, a school could improve its average by actively discouraging its weaker students from taking the SAT. This strategy only makes sense if increasing test scores rather than our children’s best interest is the goal, and these two interests can indeed be in opposition. In the perverse logic of testing, teachers and guidance councilors who are especially good at convincing their students to have high educational goals or to at least consider taking the test now just in case they may change their mind about college in the future are having a negative impact on their school’s test scores.

At this point one might be tempted to point out the No Child Left Behind mandated testing has taken such problems of comparing school to school with different populations into account. Schools are instead compared with themselves and asked to demonstrate improvement on standardized test scores from year to year. The impressive gains in standardized test scores demonstrated by many schools are cited as evidence of the success of the testing fad. The tests themselves are used to justify the testing. 

But let’s not lose sight of the goal here. The corporate model--the Lex Luthor solution--that many would like to see imposed on education has a clear bottom line, namely profits. Test scores are offered as the appropriate analogue to earnings when it comes to school performance. Just as companies can show short term profits at the expense of long term gains, schools can increase test scores at the expense of education.

Consider the following thought experiment to understand how little impressive-looking gains in test scores may be worth. You send your son to a 30 hour SAT prep course, and afterward he increases his score by 100 points (the Princeton Review in fact guarantees that the student’s score will improve by 150 points after taking such a course which amounts to as much as a fifty percentile leap). Would you now believe that your son with his dramatically increased test score is generally better educated in some significant sense than he was before completed the 30 hours of test preparation? Certainly not. Then why would we make anything more of an improvement in the average test score of a school? 

Such companies as the Princeton Review are in the business of raising test scores, not in educating our youth or even preparing them for further academic study in college. There is an important difference here. The test score business is not the same at the education business.

Waiting for Lex Luther

If our education system is truly broken as is so often claimed, then we can ignore the Hippocratic maxim “first do no harm.” Doing something--anything at all--will be at least no worse than the status quo. Of course we should always strive to improve public education, but if public education is generally working, if the statistical measures that predict our doom are bunk (and I will argue in later posts that we have good reason to think that at least many of them are), then we have something to lose in considering sweeping education reform proposals, something quite precious. 

If so, we ought to be cautious about imposing “magic bullet” reforms such as those propagandized in the 2010 documentary “Waiting for Superman,” specifically charter schools and union-busting. It was the villain, Lex Luthor, who was the industrialist and mad scientist trying to impose his vision on the world to fit his schemes. It seems to be he likes of him rather than Superman to whom too many are looking for help in improving our system of education.

The notion of “America’s failing schools” or our “broken educational system” is a cherished dogma that few seem incentivized to question. Both Democrats and Republicans find it expedient for pursuing their ends. Democrats use it for seeking greater funding for schools which helps them maintain support from teachers unions. Republicans find it just as useful for fomenting opposition to unions and painting pictures of big government boondoggles to justify tax cuts. Education theoreticians depend on the myth of failing schools to help promote their various reform proposals. Even Richard Dawkins and Sam Harris celebrated for their skepticism of dogmas are quick to attribute America’s “failing schools” to those rare nutty school boards pushing intelligent design. They don’t have an ounce of their tauted skepticism to direct toward the “failing schools” slogan itself. It seems to be one-size-fits-all evidence for support of just about any agenda. The political echo chamber reverberates with the truism that American students rank among the worst educated in math and science when compared to industrialized countries of the world, but where does this notion come from, and is it true?

First consider that claims that the sky is falling when it comes to the preparation of youth to take the lead in the future are nothing new. Growing up in the 1980’s, the big fear in my youth was that the Japanese superior work ethic in schools and factories would drive us to second rate nation status. In the widely heralded 1983 report “A Nation at Risk,” author James J. Harvey summarized the sentiments of the commission members with these memorable opening lines: "the educational foundations of our society are presently being eroded by a rising tide of mediocrity that threatens our very future as a Nation and a people.” It went on to claim that “If an unfriendly foreign power had attempted to impose on America the mediocre educational performance that exists today, we might well have viewed it as an act of war." These words will certainly sound familiar to anyone in education who is at least as old as I am (40 years old with 16 years teaching experience). 

A generation after the report, these claims of mediocre education spelling the doom of American society in retrospect are revealed as no more than the typical complaints that every generation in history has had about the moral and intellectual failings of it’s youth. These are the complaints that Socrates had about the youth of ancient Athens, that our grandparents had about our parent’s generation and the ones that our grandchildren will no doubt someday have about their children.

If “sky is falling” predictions were so wrong 28 ago, what should make us think that they are any more likely to be right when such modern seers forecast decades into the future today? The Communists haven’t invaded. Japanese men in business suits haven’t taken over the world. (Our collective fears on that score are today directed at China and India) That doesn’t mean that our future will necessarily be rosy. Our own banker’s lack of ability to understand the risks they were taking surprisingly turned out to constitute a far greater threat to our economic well-being than any foreign power in recent years. We will likely be surprised over and over again by the future. One thing we should always expect is the unexpected. We should therefore have a lot more humility about our forecasts and recognize that we just aren’t very good at predicting such things. We should try to improve education without lending any credence towards doomsday predictors that may lead us to hasty and drastic measures with perhaps dire unforeseen consequences.

Epistemological Metaphors Part 2

Peirce’s Cable, Quine’s Web, and Neurath’s Boat 
Descartes was inspired by Euclid’s geometric proofs, which demonstrated their claims to truth through incremental steps each with a simple justification in a long chain of reasoning. He likened himself to an architect needing “to demolish everything completely and start again right from the foundations,” and he sought then to build an edifice of knowledge upon such metaphysical first principles using Euclidean chains of reasoning. 

Letting the “foundation” metaphor alone for now, Peirce pointed out a key problem with the chain metaphor for knowledge. If Descartes had come up with a “long chain of reasoning” which could, by its end convince us that, say, making a cup of coffee is what we ought to do about now, that conclusion would be as weak as the weakest link of reasoning that aided in the construction of the chain. Even with a rock solid foundation on which to base our beliefs, in Descartes architectural model for knowledge, the strength of our foundation is no more important than the strength of each individual link in a chain of reasoning. Rather than model philosophy on mathematics, Peirce recommended that,

“Philosophy ought to imitate the successful sciences in its methods, so far as to proceed only from tangible premises which can be subjected to careful scrutiny, and to trust rather to the multitude and variety of its arguments than to the conclusiveness of any one. Its reasoning should not form a chain which is no stronger than its weakest link, but a cable whose fibers may be ever so slender, provided they are sufficiently numerous and intimately connected.”

Even better than having one convincing line of thought that leads to a conclusion is the possibility of generating multiple such lines of reasoning to support it. What is emphasized in Peirce’s cable metaphor is the importance of all the inferential connections between fibers or beliefs that don’t necessarily all follow in a single sequence of deductions but rather wrap around one another in mutual support.

    

The problem with Descartes’ chain metaphor is that the strength of the chain fails as easily as a single link can break. And we know that they can break since we’ve all made deductions that we thought were rock solid in the past but were later shown to have serious flaws. Quite heavy links may be found to weigh on the support of a single loop of hair, while a cable can endure the failure of a few fibers so long as there are enough other fibers remaining. It is the interconnectivity of propositions rather than a single sequence of deductions that makes a cable of beliefs strong.

Peirce’s cable could be seen as a hemp rope where no single fiber could extend along the entire length, yet, pulling as hard as we might, it won’t come apart. No single belief imagined as a fiber in such a cable or rope is any more a core or foundational belief without which the whole structure would collapse. W. V. O. Quine saw this sort of interconnectedness and mutual support for our systems of belief as a web. Perhaps we can even mix metaphors and think of the structure of our beliefs as a web of ropes or cables rather than a web of distinct fibers. In any case, we can imagine reweaving parts of such a web while the greater structure remains intact. With enough reweaving, it is possible to imagine that after a time the entire web no longer has any of the same fibers as it did before. It could in principle be completely rewoven without needing a “doubt everything” clearing away of all past knowledge; yet, at no time would the majority of beliefs in the web have gone unsupported by other strands. Our existing beliefs are all the foundation we need to extend and improve our webs of belief without any limits on the degree to which our beliefs could be improved.

While cables, ropes, and webs are excellent antidotes to chains and the myth of core beliefs, my favorite antidote to foundationalism—the thought that we need a rock solid foundation to claim any knowledge which often takes the form of a Platonic essence—is Neurath’s boat. While Descartes modeled his philosophical method on mathematics, following Peirce, Otto Neurath also took science to be the proper model for philosophical thinking. As Quine explained, rather than clearing away all but a single bedrock belief or two upon which to build a system of thought,

"Neurath has likened science to a boat which, if we are to rebuild it, we must rebuild plank by plank while staying afloat in it. The philosopher and the scientist are in the same boat. Our boat stays afloat because at each alteration we keep the bulk of it intact as a going concern."

It doesn’t merely stay afloat. In fact, as contemporary philosopher of religion and self-described pragmatist Jefferey Stout put it, Neurath’s boat

“can travel the open seas, trade with foreign places, and send parties in search of virgin timber. Its crew can take unimagined treasures on board, plunder shipwrecks for usable gear, and invent an engine to pull weight once pulled by oar.”

Presumably at that point, we could ditch the oars. Note that there is no analogue to doing so in the chain metaphor. Every link is integral to a line of reasoning, whereas in attaining greater understanding of, say, the atom, it may be useful at some point in our educations to think of it as a round particle and later to drop that notion in favor of the Rutherford-Bohr “planetary” model, later a cloud model, and later still some better not yet imagined conception that we perhaps could have never created without such scaffolding. What this ship cannot do is sail into dry dock for a complete overhaul. In Neurath’s words,

"We are like sailors who on the open sea must reconstruct their ship but are never able to start afresh from the bottom. Where a beam is taken away a new one must at once be put there, and for this the rest of the ship is used as support. In this way, by using the old beams and driftwood the ship can be shaped entirely anew, but only by gradual reconstruction."

Just as the web of beliefs can be rewoven, the ship can be rebuilt, but we can only do it from aboard the ship. Such is the human condition. This rebuilding is the process of criticizing and revising our beliefs. While we can’t call all of our beliefs into question at once—where would we stand to do that?—there is no belief we have that cannot be subjected to such criticism. The complete overhaul (question everything!) of knowledge that Descartes wanted to do is not possible in this analogy, and it isn’t possible in practice either. Though many people hoped at one time that all our knowledge could be erected as a tower on top of solid ground, such a model does not agree with how knowledge functions in practice. And just what is it, a pragmatist wonders, that we wanted to be able to do in that tower that can’t be done in Neurath’s boat?

Peirce’s Cable, Quine’s Web, and Neurath’s Boat are metaphors better suited to how knowledge acquisition functions than are Descartes’ foundation and chain metaphors. The position that Descartes imagined himself to be in deducing his existence was never the position we were actually in. We always already have beliefs. What we need to do is decide which ones to keep, which ones to drop, which ones to change, and which ones to add in an ongoing process that seems very little indeed like clearing a plot of land to erect a building from the ground up. There is no single belief such that if it needed to be updated in light of new experience it would cause the entire tower of knowledge to collapse. The reweaving of a web or the repairs and alterations on a boat while at sea seem much more to the point.

Schema Theory and The Copy View

The following is a short pragmatic take on the history of Western philosophy with regard to mind conceived of as a mirror of nature.

The Copy View
The “copy view” is the position that perception, memory, and other functions of mind are matters of finding or making mental models in the form of internal subjective replicas of external objective reality. Cognition then is the process of using these structures to draw conclusions. Knowledge in this view is having mental structures that are accurate reflections of reality. The copy view differs from other hypotheses about mental structures such as scripts, categories, and schemata in that it goes further in not merely asserting the utility of abstract notions of organizational frameworks but also holding that these representational structures have the same sort of structure of the objective reality they represent. In the copy view, the concept “automobile” is a mental version of a physical automobile where attributes of the mental copy can be matched up with attributes of the objective automobile in a one-to-one correspondence.

Platonism
The copy view of mind has roots going back at least to Plato. In the Platonic view we have knowledge if our forms (categories, concepts, schemes) are the ideal forms. We have true knowledge of “automobile” if our mental version of automobile is the ideal automobile.  In the Platonic view, even numbers share in the essence Duality, and a sick person is participating in the essence Fever. A golden retriever we encounter on the street is just a shadow on the wall of Plato’s Cave cast by the ideal essence of Dogness which is more real than any individual dog.

These ideal forms are educed (brought out from within) through proper education rather than induced (imposed from without). The ideal forms are innate. Plato thought that they are already possessed and can be demonstrated to be so by educing them through Socratic questioning.

Today we deride his tendency to treat abstract constructs as concrete entities with independent existence as “reification.” While the notion of ideal forms sounds pretty nutty today, the Platonic view of knowledge lives on to the extent we formulate mind in terms of subjective mental copies of objective phenomena. (And though we don’t generally reify Duality, or Fever, or Dogness, many people still reify some concepts as in the Christian’s maxim “Jesus is the Truth, the Life, and the Way.”)

Rationalism
Both the Rationalists and Empiricists tacitly accepted the copy view but had different ideas about how we generate conclusions from these mental copies.

The Rationalists (such as Descartes) thought that we have an innate faculty called Reason (a Platonic essence) which we can access to make proper conclusions. In fact, he thought we could generate knowledge through thinking alone (as in his deduction and subsequent momentous pronouncement that he exists, e.g. cogito ergo sum).

Empiricism
The empiricists (notably Locke and Hume) denied that there is knowledge independent of experience. They asserted that there is nothing in the mind that was not first in the senses. The is no faculty of Reason (actually, I think for Locke there was, but for the later more thorough-going empiricist Hume, no) or any knowledge whatsoever that is there in the mind until it is gained through interaction with the world. For empiricists, all knowledge comes from experience, and what we experience are our sensory perceptions. We accumulate mental copies and learn associations between them. (Thorndike was in the empiricist tradition as an associationist. Skinner does not seem to take the copy view and did not hypothesize abstract mental structures of any kind for knowledge).

Empiricists and Rationalists are historically in agreement about the premise of mental structures and knowledge as fundamentally representational in nature—the copy view. The disagreements were about whether or not there is non-experiential knowledge and whether reason is an innate faculty.

The Kantian Synthesis
Enter Kant to try to resolve the impasse between Rationalists and Empiricists. Kant thought that the empiricists had made valid criticisms of the rationalist view (such as through Hume’s thought experiment of a baby deprived of all sensations until its 18th year. It seems reasonable to think that this 18 year old would not possess any knowledge whatsoever.) But Kant thought the empiricist’s view was inadequate since there must be some mechanism prior to experience to give structure to perceptions. Otherwise no one could ever identify an object as the same object under different lighting conditions or different angles of view as sensory perceptions are never exactly the same and always in flux. Kant famously summarized, "Concepts without percepts are empty; percepts without concepts are blind."

Unlike the empiricists, Kant didn’t think reality was directly knowable through the senses because our language and senses amount to a sort of lens which intervenes between a mental “eye” and its object. (Such ocular metaphor’s for knowledge--to “see” is to know, beliefs are “views,” memories are reviewed by “the mind’s eye,”  new understanding is “insight”--are typical of the copy view and are deeply embedded in our language.) Because of the filtering properties of the “lens,” we can’t know if our copies are accurate.

Kant agreed with the empiricists that there is nothing in the mind that was not first in the senses but made an exception for what was already in the mind from the very beginning--not any knowledge or any built-in faculty of reason but rather a system that creates structure to our experiences. He struck middle ground by proposing that certain concepts or categories (certain schemes or schemata in our terms) such as for quantity, time, and causality are innate.

Piaget can be thought of as having proven Kant wrong to some degree by showing that the schemes Kant wrote about are actually developed over time rather than in-born.

Note that the schemes or categories for Kant and Piaget are not mental copies of reality but rather more abstract mental structures; however, the copy view persisted for Kantians with regard to true beliefs and accurate memories. Kant left us with the problem of being hopelessly out of touch with reality since in the Kantian picture, our senses always intervene between the mind and what it perceives and our languages impose categories on thought and limits on what thoughts are thinkable.

We don’t have a way to directly compare a belief to reality to see if it properly represents it, since in the Kantian view, reality is only indirectly accessible to mind. Only the representations (categories, schemes) of reality rather than reality itself are directly known.  Note the parallel to the Platonic problem of living in a world of shadows. 

Pragmatism
The pragmatists (Peirce, James, Dewey) got us around the Platonic appearance-reality quandary by questioning whether beliefs are best thought of as representations of reality. With the benefit of the Darwinian perspective, they saw language use and rational thought in general as evolving to cope with reality rather than evolving to represent reality.

There is no evolutionary reason why the structure of thought would need to be fundamentally representational to be useful for helping us get what we want, so it is doubtful that the “copy view” captures anything fundamental about cognition.  

While there may be some mental activity best imagined as a matter of making copies, the accuracy of copying is only of secondary importance to utility. In general, thinking is a sort of tool use, and tools don’t need to copy reality to be useful. The relevant test of a hammer is whether it drives nails, not whether it correctly copies reality. So it is with the human tools we call language and rational thought where the relevant test is the ability to coordinate human behavior to meet our various desires and needs.

Schema Theory and The Copy View
Like copies, schemata are mental structures, but their organization is more abstract. The “automobile” concept is not though of as aimed at a Platonic ideal form but a malleable framework of abstract structures. In the schema view, mind is a structure of structures as well as the on-going processes of structuring and restructuring.

In schema theory, memory is not a matter of re-presenting a stored mental copy to the mind’s eye. Memory retrieval is a process of construction.

Epistemological Metaphors Part I

In a 2007 debate between Richard Dawkins and Oxford colleague John Lennox, Lennox espoused the view that atheists can be moral but that only believers can have a foundation for morality. Atheists can of course be good, in Lennox’s view, but only a theist can tell us why we ought to be good by appealing to a metaphysical foundation for moral beliefs. In fact, a good atheist, it is claimed, can only be good because of tacitly smuggling in theistic foundational beliefs without acknowledging them as such. Likewise, the Christian notion that Jesus is Truth itself is often put forth as a necessary foundation for scientific beliefs as well. Scientism also includes such a premise that all knowledge needs to rest on something unshakable and that this view spells the doom of scientism. Under scientism, the scientific method is put forth as an alternative foundation to counter the likes of Lennox. As explained previously and as Lennox might be quick to point out, since the scientific method itself cannot be verified through the scientific method, scientism fails to serve as the sort of foundation to compete with the Christian one that Lennox claims. However, Plato's Euthyphro Dilemma preemptively demonstrated that the Christian moral foundation has the same problem of not being able to support its own weight. 

Atheists frequently also buy into the view that in order to make knowledge claims both ethical and factual, we need to reject pragmatism in favor of realism and the correspondence theory of truth. Consider for example Sam Harris's desire to find a way to say “that stoning women for adultery is really wrong, in some absolute sense.”  He seems to be demanding the sort of foundation for knowledge that Lennox says atheists are lacking. But do we need to accept the premise that we need to always think of knowledge in terms of architecture? Granted, if we do, then it indeed seems that a failure to find some bedrock on which to construct our tower of knowledge puts our whole belief structure at risk. Our tower of knowledge--the set of all our other beliefs when thought of as derived from foundational beliefs--would be no more secure than the foundational beliefs upon which all our other beliefs rest. But is knowledge construction, as the foundation metaphor suggests, really all that much like building a tower from the ground up? 

Skepticism 

The foundation metaphor for knowledge hearkens back to the father of modern philosophy, Rene Descartes, who set a standard for knowledge as only that which is “based on a reason so strong that it can never be shaken by any stronger reason.” He distinguished knowledge from mere conviction, the state of being persuaded where there “remains some reason which might lead us to doubt.” Knowledge then must be rooted in beliefs that we simply cannot doubt.

    

Generations of philosophers followed Descartes in thinking that beliefs must be founded on unassailable first principles before we ought to be comfortable in claiming to know anything. That imperative and the fact that no such principles seem readily available have left many of us feeling very uncomfortable indeed since we can find few beliefs if any that cannot ever be doubted. This discomfort is what Richard Bernstein called “Cartesian anxiety” (a term we saw earlier). In spite of all the broad success we have had in coordinating our behavior through language use to get what we want and in disregard of our ever-increasing ability to meet our needs and improve our lives through the use of our beliefs and reasoning, Cartesian thinking still leaves many of us fearing that perhaps we still don’t actually know anything at all. We could all be bodies fooled into thinking that they are having experiences that are really just part of a simulation as in the movie The Matrix.

    

Does Descartes himself who taught us to demand such a foundation for our knowledge even meet his own demand? Consider Descartes’ thought experiment from his Second Meditation:

“I have convinced myself that there is absolutely nothing in the world, no sky, no earth, no minds, no bodies. Does it now follow that I too do not exist? No: if I convinced myself of something then I certainly existed. But there is a deceiver of supreme power and cunning who is deliberately and constantly deceiving me. In that case I too undoubtedly exist, if he is deceiving me; and let him deceive me as much as he can, he will never bring it about that I am nothing so long as I think that I am something. So after considering everything very thoroughly, I must finally conclude that this proposition, I am, I exist, is necessarily true whenever it is put forward by me or conceived in my mind.”

And there you have it. Cogito ergo sum. I think therefore I am. With his apparent uncanny knack for skepticism, it seems Descartes could perform the impressive feat of doubting even the existence of his own body, but he could not, alas, doubt that he thinks. Upon this indubitable foundational belief, Descartes carefully deduced that he must in fact exist. He then “proved” the existence of God (not worth going into) and went on to reason his way to knowledge that he has a body without merely accepting at face value the reports of his possibly unreliable senses. (His argument hinged on the fact that a benevolent God would not have given him senses that were completely unreliable.) As must be obvious, it didn’t take long for philosophers to find ways to doubt the premises and chains of reasoning that Descartes had claimed were indubitable. 

What remained long after Descartes—that which warrants his designation as the father of modern philosophy—was not an edifice of beliefs we could thereafter claim as true knowledge having met the demand for solid reason resting upon an indubitable foundation. Philosophers have been able to find fault with  Descartes’ “basis on which it seems to me that all human certainty can be founded.” What remained, instead of any specific claims to knowledge, was his systematic way of doing philosophy and the belief that we really must have such an unassailable foundation to be able to claim knowledge.

    

Sam Harris was afraid that if we all became pragmatists it would be “a recipe for End of Days chaos,” because we would all then lose the conviction that we can know anything—“about anything,” he emphasized in the throes of a particularly dramatic case of Cartesian anxiety.  Harris fears that we would all have to be extreme skeptics if we were all pragmatists. I find that claim ironic because pragmatism actually began with anti-skepticism—with C. S. Peirce’s essay “The Fixation of Belief” where he pointed a way out of this Cartesian skepticism.

    

Peirce noted Descartes’ method, doubt everything!, is not something we can ever do. Though Descartes is to be admired for questioning and defending his basic assumptions rather than taking too much for granted as so many of his predecessors had done, we can get along without the radical skeptic. We can only hold beliefs in doubt with respect to beliefs to which we remain committed. There is simply no way to follow the philosopher’s dictum, “doubt everything,” and one wonders how Descartes ever convinced himself that he had done it.

    

In Peirce’s account of inquiry, there is “real doubt,” which is the irritation that moves us to inquiry, and then there is Cartesian “fake doubt,” which need not and cannot be satisfied because it doesn’t arise out of any real dissatisfaction. The aim of inquiry is always to satisfy our real doubts rather than to achieve the perfect certainty of indubitable beliefs. Peirce was convinced that inquiry has as its aim the satisfaction of doubt because when this irritation is satisfied through the course of inquiry, inquiry ceases.

   

If inquiry is concerned with satisfying the doubts we actually have, rather than being concerned with resting on a position that simply could never be doubted, then our position is very different from the one Descartes imagined in his thought experiment. We are justified in claiming knowledge when we have considered the available evidence, arguments, and counter-arguments yet have no real doubts rather than only when we are in a position to say that our beliefs could never be doubted. All beliefs can, in principle, be doubted. What we claim to know is what we have good reason to believe and have no good reason to doubt; yet all of our beliefs can be held in doubt, and any one of our particular beliefs may in fact turn out to be false. That doesn’t mean that we can never have good reason to think that our beliefs are true. Science, for example, is quite comfortable with Peirce’s notion of fallibility—that we could be wrong. The fact that scientists are willing to hold any belief in doubt if only for the sake of argument is part of what makes science so successful at finding new and better beliefs and correcting false beliefs. At least some of our beliefs are almost certainly false, but given our broad success in life, we just can’t be as wrong about the majority of them as a Cartesian worrywart might fear.