Mastery-based Grading (for Moderates)

Last week I gave a short presentation on teaching to a group of faculty at my institution. I decided to talk about an experimental assessment method I’ve been thinking about privately for the last three-and-a-half years (ever since a post at In Socrates’ Wake stirred me from my dogmatic slumber). Next semester I’ll have my first really good opportunity to implement this method. It seemed like the sort of thing worth sharing here, even though I haven’t tested it yet.

Awkward things about traditional grading. I don’t think I’m alone in hating grading. I like reading good student work, but judging it is taxing.Students may think about their grades a lot, but very few of us got into teaching in order to pass judgment on people. And whenever I’m disappointed by an assignment, I ask myself if I taught well enough, if I made my expectations clear enough, if it would be unfair of me to judge the work harshly. Grading has various functions, though, and two obvious ones in particular: to motivate students to do the work (assuming they want good grades), and to serve as a signal of student quality to folks who weren’t in the class and didn’t assess them. Normally this is done by finding a weighted average of each student’s work over the course of the semester. Maybe later assignments count more, or assignments more central to the purpose of the course (like exams and papers) count more.

There are two drawbacks to this “average performance” model. One is that grades provide poor feedback. By the time a student has a grade, they’re stuck with it even if they improve later, and there is little external motivation for them to improve unless another assessment is coming up later that evaluates them for the same knowledge or skills. Instructors usually use other methods to provide feedback (e.g. comments), but those methods don’t motivate the way grades do. A second drawback is what might be called the “improvement problem” (described ages ago by Mark Lance at NewAPPS). Imagine Student A who performs pretty well throughout the semester. Student A may have taken similar courses before, or may be unusually sharp, or may have been better-than-usually prepared for the class. Student A gets an A–. Student B, on the other hand, struggles early in the semester but perseveres until things fall into place and by the end of the term she is outperforming Student A. Nevertheless, Student B gets a B because of her weak early performance. Something just seems a bit unfair about that.

The improvement problem

The improvement problem, illustrated.

Mastery-based assessment. There are a lot of ways to address the improvement problem. One strategy is called “mastery-based grading” or “mastery-based assessment,” inspired by martial arts (earning successive colored belts) or video games (passing successive levels). In particular, Ann Cahill and Stephen Bloch-Schulman developed an ingenious design for a critical thinking course based on the martial arts metaphor. I won’t describe it here, but you can read about it yourself (I recommend it). The problem with Cahill and Bloch-Schulman’s model is that it seems difficult to adapt to certain kinds of courses—it requires students to work individually at their own pace instead of together. The result is that students cover a topic at different times during the semester, and don’t collaborate. It sounds like a great way to teach their course, but I also want students to work with and learn from each other. (In fact I am contractually obliged to teach that way.)

Moderate mastery-based assessment. So my plan is to water down the Cahill and Bloch-Schulman model enough to get most of its benefits without its drawbacks, drawing significantly on a suggestion by Dustin Locke. I have a scaffolded sequence of main papers throughout the semester, each of which focuses on specific analytical skills and builds on the skills assessed in previous assignments:

  • Paper 1: precise description of a topic or view (200 words on the topic of the week)
  • Paper 2: exposition of an argument from the reading (300 words on the topic of the week)
  • Paper 3: Like paper 2 plus an original objection (600 words on the topic of the week)
  • Final paper: Like paper 3 plus a discussion of material from outside of class (3500 words on any topic from the semester)

Now, this sequence works for philosophy courses, but for other kinds of courses one could use a sequence of different skills, or use different kinds of assessments (e.g. tests or creative projects). None of this is particularly new. The “mastery-based” twist is to allow students to turn in the papers when they choose, and as many times as they choose. Only the last grade they receive on each assignment counts toward their final grade for the course. That way, I can hold them to high standards without being overly harsh or punishing them for being unprepared for the course, since they can try again and replace their previous grades.

I may be fooling myself, but I don’t think this plan will leave me buried alive under a mountain of grading, at least in a small-ish course. First of all, each paper but the last is quite short. I have pedagogical reasons for this—it makes the assignments seem less intimidating, and encourages students to focus on the target skill instead of panicking or getting distracted by the mechanics of writing. But also it’s easier to grade a lot of short papers with clear objectives. Second, students cannot regress in the sequence. They must take each step in the sequence seriously, moving on only when they’re satisfied. It also prevents strategic thinkers from massaging their grades and the end of the term by turning in a bunch of short papers. Third, students can only submit one paper a week. This encourages them to invest in each attempt (respecting my time as a grader). It also forces them to choose new topics for each re-attempt, so that they have to practice their analytical skills in a new context. Only for the final paper can re-attempts be revisions of earlier work.

The benefits of moderation. My hope is that this assessment model improves on the problems of the “average performance” model. Grades provide better motivating feedback since they contribute to the final grade, but can be improved if the student chooses to re-attempt the assignment. This model also softens the impact of the improvement problem (though not as well as Cahill and Bloch-Schulman’s more radical model).

And importantly for me, the model is flexible enough to work in courses on diverse topics, and varied uses of classroom time. In particular, I can reserve most of class time for lectures, group work, and discussion as in a traditional philosophy course. I can also have students work together through weekly topics on the syllabus, while they work on their analytical skills at their own pace through the written assignments.

Of course, it’s possible for an extremely driven and well-prepared student to finish all her papers within the first four weeks of the term. I think that’s fine—if she’s that over-prepared she can finish early and spend more time on her other courses. There are other reasons for her to read and come class each week. Another worry that’s been expressed to me is that this plan might contribute to grade inflation. Again, I’m not worried. I think it’ll be easier for me to hold my students to higher standards this way. If they all earn As by the end of the term, they’ll each have earned it.

Advertisements

Metadicursive Technology: Claims, Views, Arguments

This is the first post in what will be an ongoing series about what I like to call “philosophical technology” or, sometimes, “metadiscursive” or “metaconceptual technology” (since it’s not relevant only to philosophers). I said last time that the method of philosophy is just the method of inquiry, but in practice a lot of philosophy these days involves a lot of attention to the way we use words or concepts (more on that next time). Since that’s a thing that philosophers do, we need to have some conceptual resources for talking about ways of talking, or for thinking about ways of thinking. I like to refer to these resources as bits of technology to emphasize the fact that developing these resources requires some ingenuity and effort, that using them effectively involves a bit of training, and that they can be developed or improved over time. I like to emphasize that last part because, like a lot of philosophical work, progress in metadiscursive technology tends to become invisible once it’s been made.

Some bits of philosophical technology are pretty well-known to most everyone—distinctions, objections, counterexamples—though most people don’t think about them explicitly or as bits of technology for getting around. Introductory philosophy classes often cover some simple metadiscursive technology. Perhaps most commonly arguments, soundness and validity for arguments, and necessary and sufficient conditions. More specialized philosophy classes will often also cover such things as the analytic/synthetic distinction, and the difference between a priori and a posteriori or empirical knowledge (and whether and how those bits of technology are useful, or even make sense, is a subject of controversy among philosophers). Some classes will cover modality (roughly: necessity, possibility, and related notions) which comes in various forms—alethic, epistemic, practical, and others. All of these notions are examples of what I’m calling metadiscursive technology, tools that philosophers use to think clearly about the ways we reason. I won’t retread over all this well-worn territory here on Explicit Content, but I suppose I would if I were to write a big, heavy philosophy textbook or a philosophical field guide (that sounds fun, doesn’t it?). But I will talk about some other bits of technology that aren’t as commonly discussed, and that I’ve got something to say about.

To start off, though, I’m going to talk about something pretty basic: the difference between a claim, a view, and an argument. Although it’s simple it trips up a lot of students writing their first philosophy papers. And you might also think there has been progress here, since in the Greek of Plato’s time there was just one word, logos, which was used to describe each of these three things (as well as “word” and “sentence” and “speech.” But not “story.”). Because this stuff is so basic, the main body of the post may be kind of boring to anyone with some experience with philosophy, but in the interest of making things explicit I’m going to write on anyway.

First, a claim is the sort of thing that is expressed by a declarative sentence. A lot of what philosophers do is examine claims, and eventually commit themselves to affirming some of them, and denying others. (If you both affirm and deny the same claim then you’ve got a contradiction on your hands, and almost everybody thinks that’s bad.) “The mind is a nonphysical substance” and “Free will is not compatible with determinism” are claims that some philosophers have made. Claims are the sorts of things that can be true or false, or plausible or implausible. Some claims have become important enough that they get their own names. The second claim above is called incompatibilism. Another named claim is hedonism, which is the claim that “The only thing that is good for its own sake is pleasure, and the only thing that is bad for its own sake is pain.” Claims get more complicated when you consider that there can be claims about claims, and claims about arguments, and so on. For example someone might claim that “Hedonism is false,” or that “Incompatibilism is true because the consequence argument is sound.” But although consideration of claims is a necessary part of philosophy, the job of the philosopher doesn’t stop there.

Views are collections of claims that are supposed to be coherent. Views, like claims, can be true or false (though it’s more common to say they are right or wrong), and sometimes have names. Often a view is supposed to explain various things. For example, an old view called machine functionalism was supposed to explain why humans and octopuses can have the same mental states, like pain, where the view called mind-brain identity theory could not. Incompatible views may still share claims. For example, libertarianism and hard determinism are names for views about free will. Both the libertarian and the hard determinist believe in the claim that “Free will is not compatible with determinism,” but the libertarian is committed to the claim “We have free will,” and denies the claim “Determinism is true.” The hard determinist takes the opposite attitudes toward those other claims.

Views are often associated with the particular philosophers who explain them, like Ruth Garrett Millikan’s teleosemantics (roughly a view that meanings of words or thoughts, like biological functions, are determined by their causal history according to a process of natural selection). But popular views often fragment. Now teleofunctionalism is a word for a family of related views, like Millikan’s and Karen Neander’s. There is a folk caricature of philosophy that it consists in the elaboration of lots of different views, and that philosophers are people who know about lots of views and prefer some of them. But views are not the dominant currency of contemporary analytic philosophy, either.

The main business of philosophy involves giving and evaluating arguments. Arguments are what I focus on most when I teach philosophy, and there are a few different ways to describe them. The slogan that I’ve been using lately is that an argument is a reasoned defense of a claim. On this view (see what I did there?), arguments consist of two parts: a claim, called the conclusion, that the argument is supposed to support, and a reason that supports the conclusion. Philosophers use arguments to support claims (where the claim is the conclusion) and views (where the various claims that make up the view are conclusions, usually of different arguments). And just like there can be claims about claims, there can be arguments about arguments. For example, criticisms or objections about arguments are arguments about arguments (they are arguments that some other argument is bad). Opinion essays and most papers for classes are structured around a main argument (and usually contain other arguments). The conclusion of the main argument is what you call a “thesis” when you’re learning to write in school.

If arguments are reasoned defenses of claims, then you see that they are not bare statements of claims, and not disputes or questions or problems. Philosophilcal controversies, like the “mind-body problem” or the “problem of personal identity,” are not arguments in this sense because they do not have conclusions and they do not provide reasons. People make arguments for various views that resolve these controversies in different ways, but philosophers do not usually call the controversies themselves “arguments” (it gets confusing quickly).

Monty Python on arguments.

The most common way to model arguments in the analytic tradition is based on the form of a deductive inference in classical logic, or a syllogism in Aristotelian logic. Either way, the reason is made up of claims called premises that, if they are arranged right, support the conclusion through some rule or combination of rules (but rules are not just like more premises, as Lewis Carroll demonstrates in a well-known story). I say that logic models arguments, rather than saying that logic expresses arguments, since most philosophical arguments are given in the form of reasons, not in the form of logical deductions. But modeling arguments with logic can be very useful for determining whether the argument is good, and determining how an argument might be weak (undergraduates take note). Philosophers sometimes model their own arguments in logical form, and often model arguments in order to make clear objections to them.

It’s important to note that arguments cannot be true or false. Conclusions or premises, since they are also claims, can be true or false, but arguments have more complicated ways of being good or bad. Arguments can be valid or invalid, or cogent or not cogent, or sound or unsound, and so on. It would be tedious to explain what all these terms mean here (that’s taught in most introductory philosophy or logic classes, at least as they apply to deductive arguments, and perhaps inductive arguments). Basically, though, an argument is bad if it doesn’t give you a good enough reason to believe its conclusion, and the ways that reasons are bad are different and more complicated than being false.

There are at least three main ways to criticize arguments in philosophy. First, one can claim that the premises or presuppositions of the argument are untrue. That doesn’t make the argument “untrue,” and it doesn’t mean that the conclusion is false, it’s just one way that an argument might not give you reason to believe its conclusion. A second way to criticize an argument is to say that the reason doesn’t support the conclusion, regardless of whether its presuppositions are true. A simple example:

Edinburgh is in Scotland.
Humans often wear clothes.
Therefore, George Clooney is famous.

The premises and the conclusion are all true, but the premises don’t support the conclusion. They don’t give you reason to believe it. (However, this does not always make an argument invalid in classical logic… another reason to say that logic only models arguments.) A third way to criticize an argument is to claim that we have an independent reason to believe that the conclusion is false, and that this reason is better than the reason given in the argument. (But undergrads should note that it’s tricky to make a good paper out of a criticism like that unless you also figure out what specifically was wrong with the first argument, and explain it clearly.)

That’s a very brief and introductory pass at arguments, but arguments are complicated creatures and it can be difficult to understand all the ins and outs of how they work, especially when they are complicated or abstract (that’s why philosophy is hard). Or when they are about things that are difficult to think about clearly. But we deal with them all the time in life, whenever we are considering what to believe or what we should do. Whenever we consider what our reasons are, and whether they are good enough for believing or doing something, we are reflecting critically on arguments (even if we’re not doing it very clearly or explicitly). Even if not all philosophy is about arguments, critical examination of arguments is a central activity of philosophers, especially analytic philosophers. And while most disciplines do the same thing a lot of the time, philosophers are often the ones that are most concerned with developing the metadiscursive technology for doing so with self-conscious clarity and precision.

In fact, I think the skills of using basic metadiscursive technology are the most important things to teach in an introductory philosophy class, where many or most of the students do not intend to major in philosophy. Sure, it can be fun for some students to learn a bunch of views about free will or personal identity or ethics, but the most valuable and transferable skills for liberal arts students are the ones that are learned from sustained attention to arguments, and for the complicated ways of supporting and evaluating claims, arguments, and views.

The Method of Philosophy Is the Method of Inquiry

In my earlier post on the method of philosophy I made several negative claims: the method of philosophy is not based on intuitions or reflective equilibrium, it’s not random speculating, and it’s also not just about arguments. Today I’m going to motivate a little maxim that I’ve been mumbling to myself for a few years: that the method of philosophy is the method of inquiry.

What do I mean by ‘inquiry’? By ‘inquiry,’ I mean something like the deliberate project of understanding the world (including ourselves) better. Sometimes this is done in order to accomplish a specific goal, like curing polio or building bridges, and sometimes it’s not. I take it that building the Large Hadron Collider and looking for the Higgs boson is an example of the latter kind, although there have been highly practical discoveries along the way and this was always a part of the plan. At its best moments, the academy (I don’t mean the Academy, but academia, the worldwide system of universities and other institutions of higher learning) is an institution dedicated to furthering inquiry and disseminating the resulting understanding to students and others. I am tempted to think of inquiry as a distinctively human project (as far as we know). I don’t think that when a cat figures out how to use door handles it’s performing inquiry, but maybe we can say it’s a special kind of cat-inquiry as long as we recognize the differences between cat-inquiry and human inquiry. For example, the understanding gained from cat-inquiry does not tend to be disseminated among other cats, whereas human inquiry is a deeply social project.

It’s important, especially when we take into account the humanities and not just the sciences, that inquiry is about understanding of some sort, and not just truth or knowledge, narrowly construed. As I understand it, a lot of humanistic scholarship is concerned not with the articulation and justification of true claims, but of promoting a variety of ways of understanding things (texts, events, social structures). And honestly, that sort of thing goes on in science, as well. The psychological and linguistic literature, for example, are full of discussions that encourage readers (colleagues) to think about the mind or about language in a particular way. That is a lot of the force of, for example, theoretical discussions on embodied cognition, or descriptivism in linguistics.

(Another story I like to tell is that “philosophy” comes from the Greek, meaning love of wisdom, but that a lot of analytic philosophy seems rhetorically fixated on knowledge, scientific and otherwise, rather than some broader conception of human intellectual capacities.)

I believe that the method of philosophy is just the method of inquiry—that the acceptable methods in philosophical work are any and all of the acceptable methods in inquiry in general. To illustrate what I mean, I’ll talk a little bit about philosophy as a scientific discipline, and then about philosophy as a humanistic or perennial discipline.

Aperture Science: Safety First!

The method of science. Image: Valve

Philosophy and scientific inquiry. Despite what you may have heard, philosophy and science are pretty tight. This is true in at least two ways. For one, a lot of contemporary analytic philosophy draws on empirical premises to make arguments. Hilary Kornblith is a good avatar for this practice in epistemology. Kornblith uses results in psychology and cognitive science to defend a particular picture of how we come to know things, and what our limitations are. This is also very true in my own specialty, philosophy of cognitive science. For example, Andy Clark draws on a variety of scientific results to argue for a picture of how humans and other cognitive subjects are embedded in the world (review), and Jesse Prinz draws on a lot of literature to argue for a particular view of emotion (review). Some philosophers even run their own experiments. So this kind of philosophy is not, whatever its faults may be, out of touch with science, and it does not claim to produce a priori knowledge (knowledge we can acquire without experience). But there is a second way that philosophy and science are closely related: scientific inquiry is based on philosophical methods like argument and conceptual clarification. The heart of any scientific article is an argument where the results of experiments or studies are presented as premises. The construction of scientific theories and models involves systematizing information from such arguments and refining the concepts of the theory so that knowledge can be represented perspicuously. In fact, you could say that scientific methods are a special case of philosophical methods.

Now, I am not claiming here that philosophy is better than scientific disciplines. I am just saying that, at bottom, we are all doing the same kind of thing. Craig Skinner (an interesting fellow) argued online last year that one function of philosophy as a discipline is to be a source for the ‘budding off’ of other disciplines, like the sciences. This is an interesting notion, but the ‘budding off’ activity makes more sense if there is at bottom a continuity between philosophical inquiry and other kinds of inquiry.

The tree of knowledge flowers and fades.

The perennial aspect of philosophy. A lot of contemporary analytic philosophy is on board with scientific inquiry, broadly construed. It sees itself as dedicated to the project of uncovering and justifying new knowledge and understanding through argument and reasoned speculation, just as science does. However, there is another dimension to some philosophy, which I am tempted to call the “perennial” side to philosophy (with apologies to Leibniz and Sellars and others, who meant different things by “perennial philosophy.” They meant philosophy about God and other eternal things, but I mean the philosophy that comes back year after year). Philosophy in its perennial mode engages with topics that are not suited to being settled once and for all, but that require repeated engagement. I think some ethics is like this—society changes, which means that the same ethical principles need to be worked out and again and again, but in new circumstances. This gives rise to topics like medical ethics, nanoethics, and ethical debate over privacy issues in the world of big data. I know less about humanistic inquiry than I do about science, but I suspect that this sort of effort to keep our understanding up to date with new developments, and to update our understanding of old developments, is of a piece with scholarship in the humanities. And there is an important perennial dimension to philosophy instruction. Even if no progress were made in philosophical research, I think it would valuable to help successive crops of students to question Cartesian dualism and free will libertarianism.

I also think that this is the sort of picture of philosophy that the later Wittgenstein had in mind when he defended his “therapeutic” conception of philosophy (the most famous remarks are probably §§115–128). Wittgenstein claimed that philosophy clears up the linguistic confusions that we encounter in life. I wouldn’t go so far as Wittgenstein here, but I think it is a part of philosophical inquiry to devise methods for getting around logical space without getting lost, and techniques for finding our way if we have. The perennial vision of philosophy is also championed by Richard Rorty. In a notorious discussion of Derrida’s work (“Philosophy as a Kind of Writing”), he criticizes the analytic philosopher’s “Kantian” conception of inquiry as narrowly knowledge-producing, and suggests instead that philosophers think of themselves as commentators in a great and interminable conversation about how to live.

Now, thinkers like Wittgenstein and Rorty have earned their share of stigma from more mainstream analytic philosophers. There are plenty in the analytic tradition who are uncomfortable with the perennial function of any inquiry, let alone of their own discipline, and respond by writing off their perennially-oriented peers as non-philosophers, or as not truly a part of the analytic tradition. There are several figures, especially at the University of Pittsburgh, the University of Chicago, and Harvard, who get this treatment. I am not claiming here that Wittgenstein and Rorty are right about philosophy. After all, my own research is of the more scientific kind described above (though it is Wittgensteinian in other ways). But I do think that they are not wrong about what they and many others do, and I think it’s cheap to reply by insisting that what they do is not really philosophy.

Bradley Garrett, self-portrait atop the Forth Rail Bridge

The synoptic vision of Sellars’ philosopher. Image: Bradley Garrett.

So what is philosophy? All this is to say that philosophy’s methods include those of other areas of inquiry. Sometimes philosophy uses the scientific method. Sometimes science uses philosophical methods. Sometimes philosophy functions to apply old views to new situations. Sometimes in philosophy we reimagine the old and familiar from a new perspective. So if there is any method to philosophy, I think it’s just the method of inquiry in general. Philosophers adopt a broad range of methods for understanding the world, and those methods seem to include, well, all of them.

But I think in the end this is an ecumenical conclusion (I like my conclusions ecumenical). If the project of philosophy is, at its broadest, just the project of inquiry, then that sits well with a lot of other things people have said about philosophy. It sits well with Plato’s old line that “Philosophy begins in wonder,” since the ultimate end of philosophy is to promote understanding. It also plays nice with the other claims of Plato’s Socrates, that philosophy is the means to the examined life, since a better understanding about how to live well is a special case of inquiry, and perhaps the most important one. My conclusion explains the central importance of argument to Plato’s Socrates, but also the importance of critical examination of assumptions and the development of tools for navigating logical space. My conclusion also sits well with Skinner’s suggestion that philosophy is a source for other disciplines to “bud off” from, since other disciplines represent more specialized approaches to inquiry.

Finally, I think my conclusions is a happy companion to Sellars’ famous dictum that “The aim of philosophy, abstractly formulated, is to understand how things, in the broadest possible sense of the term, hang together, in the broadest possible sense of the term.” Sellars continues that philosophy is distinguished from the special disciplines in that philosophers aim to keep track of the big picture (“It is therefore the ‘eye on the whole’ which distinguishes the philosophical enterprise”). I would say that that’s a nice regulative ideal, and it is something I try to do, but I doubt it’s a necessary condition. A lot of philosophers specialize quite narrowly, who are still philosophers. And people of many professions sometimes address philosophical concerns without necessarily reflecting on how their “bailiwick fits into the countryside as a whole.”

If this is right, though, there is a question left outstanding. What does that mean for the subject matter of philosophy? Surely if “leftovers” is too narrow a characterization, then “everything” is too broad! Not everything is philosophy, but I am not sure how to limit the subject matter because anything could be philosophy. Consider that Plato’s prescriptions about policy and social architecture belong to philosophy. Aristotle’s early biology is philosophy (even if it’s not great). Newton was a natural philosopher. Really, it’s one of the most frustrating things about philosophy that potentially anything can be relevant to anything. As a philosopher of cognitive science, I feel like I should know so much more than a human being ever could. I’ve got to manage my time and effort, of course, and it’s hubris to think that one person can be an expert in everything, or even (these days) in very many things at all. But I don’t think I can write off any sphere of human knowledge as clearly irrelevant to philosophy or even to my project. I never really get to say “That’s work for another department,” unless I mean that I just don’t have the skills or the time or the funding to look into it. But I never meant to claim that it’s my job to know everything (what a wonderful and terrible job that would be!). But any technique that anybody uses to understand the world better is a technique I could potentially find a use for in my line of work.

Philosophy Isn’t All about Arguments

One of the most obvious questions one can ask about philosophical methodology is “Well, what is the method of philosophy?” If you’ve got an answer to that question in your pocket, it will help you to judge whether something is a bit of philosophy or not, and whether a bit of philosophy is a good one or a bad one. By comparison, one might suggest at a first pass that the method of science is essentially empirical: you have a question about what the world is like, and then you go check the world with a controlled experiment and find out. (That’s a potentially dangerous simplification, but it is a common thought and dealing with it is a topic for another day.) A similar caricature of the philosophical method is that a philosopher, ideally an old white dude with a long beard, sits in an overstuffed armchair stroking his facial hair. He takes his ‘intuitions’ about cases as data, and tries to make them systematic (perhaps through some kind of reflective equilibrium). That might describe the method certain patches of philosophy (except, not the part where the philosopher has to be white or male or bearded, or the part where it’s done alone). But intuitions (e.g. about what is just, or what is moral, or what counts as knowledge) don’t play the same role in all of philosophy, particularly not in philosophy of science. Consider a famous, big question in philosophy of science: realism vs. anti-realism about science. Roughly, that dispute is about whether we should believe that things like electrons and the strong force are real, or whether they’re just useful fictions that we use for predicting observations and building bridges. Arguments about scientific realism do not usually turn on accounting for our intuitions about what is real and making them systematic and discovering whether electrons belong inside or outside; they usually rely on claims about how evidence and explanation work, and how they apply to the interpretation of scientific theories and models.

So what can we say about the philosophical method? As in so many things, you can’t go too wrong starting with Plato. Plato has Socrates say somewhere that “Philosophy begins in wonder.” A lot of people seem to like that expression, and it might be true. But—my weird and enduring love for Plato notwithstanding—it doesn’t do much for me. A more articulate suggestion in Plato is that philosophy (sometimes ‘dialectic’) is the ‘examination’ part of the examined life. It is the investigation of your reasons for thinking what you think and for doing what you do, and the policy of offering those reasons up for criticism by others. I like this a little more, and it’s the kind of story I tend to take into the classroom when I teach. Generally speaking, I think Plato’s dialogues are full of great lessons about philosophical methodology. But this Platonic story is very personal. It’s about how philosophy fits into an individual life. If that were what philosophy as an academic discipline were about, it would mean that philosophers are ipso facto stupendously self-involved, or nosy, or both. Now, an enterprise whereby the participants investigate their reasons and those of their peers sounds like an interesting club, but it’s not a story about a respectable academic discipline. An academic discipline should do something for people other than those who participate in it.

A nearby suggestion, though, is that philosophy is about evaluating reasons as such. I once heard a story from Professor James Shaw that he sometimes tells his undergraduate classes. I like this story. Shaw says that the method of philosophy is described by something called the “science of argumentation,” which is presumably a generic variation on formal logic. (“Argument” here, as in most philosophical contexts, means a reasoned defense of a claim, not a verbal fight.) On Shaw’s suggestion as I understand it, philosophical training involves acquiring special knowledge of the forms of argumentation, with a focus on which ones are conducive to preserving truth, and expertise in clarifying and evaluating arguments as such. That’s the method. The content of philosophy is what he calls “leftovers.” On this view philosophy is about whatever is left unclaimed by the other disciplines (this can’t be all that philosophy is and Shaw knows that perfectly well, but it’s still a cute thought). “Leftovers” sound unsavory at first but they’re not, because there are a lot of interesting topics that can’t be adequately treated by the methods of other disciplines—Shaw’s examples were the mind-body problem, the problem of free will, the problem of personal identity, and whether death is harmful. The reason philosophy gets all the leftovers is that if you don’t have more specialized ways to discover knowledge on a topic, the science of argumentation is all you’ve got left!

I really like Shaw’s story, although I have two worries about it: philosophy isn’t just about leftovers, and its method isn’t just the science of argumentation. (But don’t tell my students I said that.) Philosophy can’t be leftovers because a lot of philosophy covers topics that are well covered by other disciplines: philosophers study language (especially semantics), but that’s now the turf of linguists as well. Philosophers butt heads with neuroscientists and social scientists about the will, responsibility, politics, epistemology, memory, concepts, and other topics. Philosophers get to keep the mind-body problem, probably, but frankly the computer scientists think that they do a better job and the philosophers have dropped the ball on that topic and let it languish since the 1970s. Moreover, there are the subdisciplines of philosophy called the philosophies of the sciences: subdisciplines like philosophy of physics, philosophy of biology, philosophy of nanoscience and, of course, philosophy of cognitive science. In all of these subdisciplines, philosophy takes on subject matter that is precisely the subject matter of other disciplines.

Concerning the method of philosophy, I’m pretty sure that it’s something like the standard line in analytic philosophy (the tradition in which I’m trained) that the method consists in attention to argument. Speculation without argument shouldn’t be dignified with the name ‘philosophy.’ In contrast to the standard line, there seems to be a view among the layfolk that philosophy consists in unconstrained speculation about abstract questions. On this view, the general form of a philosophical discussion is something like this:

JONES: “Hey, dude, what about if X?”
SMITH: “Sure man. But wait, what about if Y though?”

The thought is that philosophy consists in producing random suggestions, and philosophical expertise involves knowing what all those suggestions have been. Maybe this is why, when you tell someone at a party or on an airplane that you study philosophy, you quickly get the question “Which philosopher do you follow?” When it comes to teaching, I try to make a little show of insisting on the standard line, and disabusing my students of this layfolk cartoon. I think that this policy pushes my undergrads the right way. But I also think that the standard line goes too far, and that there is something right about the cartoon picture.

Even in analytic philosophy, good work does a lot of things apart from describing argument. For example, good work sometimes describes the range of possible ways of thinking about a topic. As we sometimes say, it maps out the “logical space.” This involves making distinctions, which is a crucial piece of philosophical technology. The standard line would hold that making distinctions is useful only insofar as it cleans up your definitions, and thereby improves an argument that features those definitions. It is, after all, implausible that argumentation is all there is to philosophy; there are at least other things one must do in order to argue effectively. For example, understanding arguments in texts, defining one’s terms, and anticipating counterarguments. What the standard line holds is not that argumentation is all there is to philosophy, but that argumentation is fundamental. All other activities belong to philosophy insofar as they help with the activity of articulating good arguments. But I’ve often walked away from philosophical discussions or from reading a paper with a distinction that was more memorable than the argument. Now, distinctions are wonderful pieces of philosophical technology. After all, the ability to make the right distinctions is a conceptual skill that you can take into new puzzles and new situations. Distinctions and other tools for navigating “logical space” without getting lost or overwhelmed are often more widely applicable than a grasp of particular arguments and counterarguments. For example, one set of distinctions familiar to most who have taken introductory philosophy is the standard tree for categorizing views about free will (below). That tree gives you a way of beginning to carve up the space of views about free will, and allows you to distinguish between two very different kinds of view that allow for free will (libertarianism and compatibilism). I would guess that most students remember that tree better than they remember, say, the consequence argument.

standard views about free will

Another activity of philosophers, and one that is harnessed by the folk picture, is the articulation of possibilities that have not been thought of or put clearly before. In particular, I mean identifying unnoticed assumptions and discarding them. There are complications here… at some level of generality, there is probably nothing new under the sun anymore (sorry, undergrad essay writers, I just about promise you didn’t think of it first). And novelty is not interesting for its own sake, anyway. But I’ve heard a few people say that one of the most important activities of the humanities and social sciences in general is to “unmask ideologies,” which I think is a way of talking about identifying and often discarding tacit assumptions. According to Robert Brandom, this is more or less the basic shape of Hegel’s method and the form of most of John McDowell’s discussions. McDowell can probably be accommodated straightforwardly by the standard line, since his writing is structured around arguments, but the writing of Hegel and of many continental thinkers is a tougher case. If one insists that everything in philosophy must be in the service of an argument, then those writers will be considered egregiously inarticulate (that may be a just criticism of Hegel, but not of e.g. Nietzsche). And it’s petulant and problematic to simply insist that they are not true philosophers.

This exploratory side of philosophical activity is easy to miss in the analytic tradition because most papers are organized around arguments, even when they include other kinds of intellectual work. However, that’s not universally true even in the analytic tradition. For example, it is possible to find arguments in Wittgenstein, but it’s difficult in most of the Tractatus, and many of the most memorable parts of his later work are not well-described as arguments. Wittgenstein’s fans in philosophy often quote him like scripture: “The world is the totality of facts, not of things”; “Nothing is hidden”; “Light dawns gradually over the whole.” This tendency infuriates some of Wittgenstein’s critics, who see him as a crackpot guru and not a philosopher (“Where are the arguments?”). I would venture that a lot of the power of Wittgenstein’s work comes from his ability to get you to think about things a certain way. Perhaps the same can be said of many of the famous thinkers in the continental tradition. I might say that Plato has a similar effect, even though his dialogues are full of arguments (after all, many of them are unconvincing on their face). A lot of the most compelling bits in Plato are the myths and stories he tells—the cave in Republic, the chariot in Phaedrus, or Aristophanes’ myth of the origin of love in Symposium. But the truth is that even though Wittgenstein and Plato have plenty of fans in the analytic tradition, analytic philosophers often avoid admitting that they are compelled by anything but sound argumentation.

So what is the method of philosophy? My opinion is that the method of philosophy just is the method of inquiry, but explaining that claim and arguing for it would require more space than I want to take up today. I’ll stop for now with a negative conclusion: attention to argument is extremely important in philosophy, but I don’t think that argumentation is the foundational activity of philosophy.