Philosophy, Film, Politics, Etc.

Friday, June 29, 2012

Theological Noncognitivism, Redux

I want to consider the idea that there are three different common uses for each the following sentences:

(1) God does not exist.
(2) God exists. 
I will outline a version of theological noncognitivism (TN), which is normally taken to entail the belief that (1) and (2) are neither true nor false.  The idea of noncognitivism in general (be it theological, moral, or what have you) is that the concepts in question are not propositional, which is another way of saying that they cannot be evaluated as either true or false.  TN implies that (1) and (2) are non-truth-evaluable under religious usages of (1) and (2).  However, what I will argue is that TN can acknowledge truth-evaluable versions of (1) and (2), and that noncognitivists can even assent to common uses of (1) which are truth-evaluable.

It might look like I'm trying to mix noncognitivism with cognitivism, but that's not entirely true.  It also might look like I'm trying to tweak noncognitivism to make it more adaptable to atheism, but that's definitely not true.  I'm trying to find the best way to understand how people talk, theists and atheists alike.  I think a lot of atheists, and some of the most prominent ones, might be noncognitivists in just the way I am about to describe.

The three ways of using (1) and (2) might all raise interesting philosophical and psychological questions, but I'll try to keep it simple.  Let's call the first way the mythological way.  In this way, people use (1) to truthfully assert of a mythological character that it is not a real entity.  Thus, we might say that Zeus does not exist, and only mean that Zeus is a character of mythology and not a real intelligent agent acting in the world.

The first point I want to emphasize is that, in using (1) in this sense, we do not have to have a coherent definition of God as an intelligent agent acting in the world.  The point is rather that God is defined as a mythological creation.  This is a coherent definition with cognitive content.  For example, we can analyze Zeus as a mythological character, and we can discuss and analyze the implications and importance of Zeus in many ways.  So, "Zeus does not exist" does not require anything more than a stipulation that "Zeus" is the name of a fictional character.  Thus, if we take the God of Christianity as a mythological character, we can say (1) and only stipulate that "God" is the name of a fictional entity, and not a real intelligent agent.

The second point I want to emphasize is that the mythological usage of (1) is not a direct rejection of religious belief in God.  A person can be a theist and still claim that (1) is true under the mythological reading.  In other words, a theist can assent to (1) and (2) simultaneously, but under different readings of "God."  Of course, a theist would not normally assent to both (1) and (2) without trying to explain what they meant.

(2), when used to express a religious or theological belief, does not assert that a fictional character is real. The theist, when uttering (2), is not saying of a fictional character that it is a real intelligent agent. (That would be an obvious contradiction, since fictional characters are by definition not real intelligent agents.)  Rather, the assent of (2) under the mythological reading is the claim that the myths are depictions of some real intelligent agent, without attributing any specific characteristics to that agent.  Thus, a theist might claim that the Bible is the word of God, and that it details the reality of God, without having any conception of what "God" might denote. Perhaps this is the way children first think about biblical stories, before they develop a theological attitude towards religious language. Thus, (1) can be a direct rejection of their mythological belief.

Another way of using (1) and (2) is theological.  The theologian takes "God" to denote an existing entity and claims that God exists in the actual world.  Theists, at least of the more intellectually advanced variety, have theological tendencies. They therefore normally appeal to some definition of "God" to explain their religious belief, though I don't think most theists demand that a clear and coherent definition is available.  They are often, perhaps even most often, satisfied with what they think is their own, personal, and incommunicable understanding of (2).  They suppose that, if somebody doesn't understand it, then that person is spiritually lost.  Still, this theological use is made in an attempt to say something true.  It is an attempt to express a proposition.  The theological noncognitivist has a problem with this, as I'll explain.  But first, let's look closer at what the theist is doing.

The theist also has a religious way of using (2) in addition to the theological one.  According to TN, religious expressions of belief are noncognitive.  They do not express a proposition.  TN takes the religious use of (2) as something other than a statement which might be true or false.  So what is the theist doing when (2) is voiced in the religious mode?

The religious use of (2) produces in the speaker certain emotional or psychological attitudes towards the world which may not so easy to define. It may also produce in the properly conditioned audience similar emotional or psychological attitudes.  It may be an expression of group loyalty, or maybe that is oversimplifying it.  Generally speaking, the religious use of (2) is a noncognitive action: by claiming that God exists or that one believes in God, one presents themself in a particular way as one who is subservient to an incomprehensible authority, and one simultaneously imposes the force of that authority on others.

This action might have other effects as well.  A person can create an alibi for their inability to ground their individuality or moral sense in their experience and knowledge. This act of commanding authority via an alibi is meaningful, and that's what I take the religious use of the word "God" to do.  I'm open to the possibility of there being other psycho-social dimensions to religious language, but my supposition is that, whatever those dimensions are, they are noncognitive.  The religious use of "God" has no cognitive coherence.  It does not describe an entity which may or may not exist.  It does not circulate as a denoting term at all.

Theology is the attempt to rationally and systematically reveal or discover the nature and truth of religious belief. Theologians thus try to iron out the inconsistencies in various meditations on God and spirituality. They try to construct a consistent view of God and the world, usually in conformity with holy texts.

TN also attempts to make sense of the use of religious language. However, it is unique in that it does not stipulate that there is any religious truth to be found. It claims that religious language has another function. Thus, TN rejects cognitive forms of theology as confusing the very nature of religious language. According to TN, if a theist tries to interpret (1) or (2) in the religious sense as being either true or false, that theist is making a category error. They are making a mistake about the sorts of things (1) and (2) express in the religious mode.

Theologians have tried for ages to provide a coherent definition of "God," such that God's existence would seem self-evident, rational or even necessary.  Thus, we have a theological usage of (1) and (2) which does not faithfully represent the religious usage and which is distinct from the mythological usage.  According to TN, there is no coherent theological definition of "God" such that God could be said to exist.  In other words, the theological idea of God is not a logical possibility.  It's just incoherent.

This might be best elucidated with possible world semantics.  The basic idea here is that when we speak about what is or is not possible, we are describing possible worlds.  Possible worlds are not actual worlds.  It is a given that there is only one actual world.  Yet, if some x is conceivable, it exists in some possible world, which is to say that there is a coherent description of a world which contains x, even if that world does not exist.  With that in mind, consider two possible readings of (1) in what I call the theological sense:
(1.1) God does not exist in the actual world.
(1.2) God does not exist in all possible worlds.
According to TN, (1.2) is true, where "God" is taken in the theological sense.  Noncognitivists can deny the existence of God, disbelieve in God, and so on--all the sorts of things atheists do.  In so doing, the noncognitivist means that God is simply a logical impossibility, an inconceivable concept, which certainly does not denote anything in the actual world, since it does not denote anything in any possible world.

To summarize the TN position which I am advocating:

  • There is no coherent theological definition of "God." 
  • Theological cognitivism is an unfaithful representation of religious language.
  • We may speak of God as a mythological figure.
  • Atheism, as the rejection of religious belief in God, is justified.
It follows that theological noncognitivists can assent to (1) in three different ways.  In the first way (the mythological way), they can mean that "God" is the name of a fictional character and not a real intelligent agent.  In the second way (the theological way), they can mean that the theological definition of "God" is inconceivable, and that God therefore cannot possibly exist.  In the third way (the religious way), the noncognitivist is rejecting the religious use of (2), denying the assertion of moral authority advanced by the theist.

I think many atheists, and some prominent ones, might be open to TN as I have described it.  Richard Dawkins, for example, has argued against the probability of a certain conception of God, and he has been criticized for turning God into something subject to the laws of science.  Dawkins' reaction is to claim that, if God is outside the bounds of a scientific definition, then he thinks the whole idea is meaningless.  That is not explicitly the noncognitivism I advocate, but it might be close to it.  Daniel Dennett also seems very close to noncognitivism when he claims that religious belief is not belief proper.  The implication is that religious belief is not a propositional attitude, which would suggest noncognitivism with respect to religious belief.  Generally speaking, it is quite common to find atheists who claim that religious language is incomprehensible, that theists don't have a coherent notion of God to begin with, and that religious institutions manipulate emotions rather than engaging intellects.  It seems to me that TN is already a large part of atheism today, but perhaps one that has not fully been realized.

As a final note, I'll mention that there is room for debate about what justifies the conclusion that theological definitions of "God" are inconceivable. Some say that, if the definition of "God" is not obviously self-contradictory, then it is coherent. However, I think the meaning and coherence of terms (or the lack thereof) is not always obvious. The fact that somebody thinks an expression makes sense does not necessarily mean it makes sense. While different people tell me that various definitions of "God" make sense to them, and therefore seem to denote possible entities, I remain unconvinced. I'll discuss why in another post.

Monday, June 25, 2012

Practical Ways of Thinking

Jason Stanley says that we must understand the need for practical ways of thinking if we are going to understand what it means to know how to do something.  His thesis (Know How, 2011, p. 130) is this:  "What happens when I acquire skill in the activity of catching fly balls?  What happens is that I come to the realization that a certain way of catching a fly ball, which I think of practically, is a way that will give me counterfactual success in fly ball catching."  Thus, one must not simply have a properly practical way of thinking about catching a fly ball; one must also realize that they have this practical way of thinking about that way.  One must thereby come to the conclusion that the way is a good way of doing it.

Stanley's account implies two ways of thinking:  The first way is the practical one, and the second is the way of thinking about the practical way of thinking such that it is about a good way of acting.  Is this second way of thinking also practical?  It must be contemplative or analytical, as it requires assessing the merits of one's practical way of thinking.  It does not seem to be action-based in the way that practical ways of thinking are supposed to be action-based.  That is, one does not employ one's evaluative way of thinking about one's practice whenever one practices.  When one engages in intelligent practice, one employs the practical way of thinking without reflectively thinking about it as such.  If one is employing knowledge, then, it is of a decidedly practical kind and does not require the other, intellectual kind of knowledge.  Yet, Stanley says our know-how involves knowing that our way is a good way.  It is not just practical knowledge.

If I am properly understanding Stanley's account, it is not intuitive. If we have a practical way of thinking about catching a fly ball and if we utilize this way of thinking in the catching of fly balls, then why couldn't we say that we therefore know how to catch fly balls?  Why must we have a separate act of realizing that we have this practical way of thinking?  Isn't the know how evident in the way of thinking itself, and not in the realization that one has a good way of thinking?  Doesn't one manifest the practical knowledge, and not the intellectual knowledge, when one intelligently catches a fly ball?

Stanley's analysis requires that the same ways of thinking we employ in thinking about our performance are integral aspects of our performance.  This is obvious from the fact that he argues for the existence of practical ways of thinking by observing the ways we think about our performance.  According to Stanley's analysis, practical ways of thinking are evidenced by reflection in ways which suggest proprioception or introspection.  He writes:  "It is simple to imagine cases in which one thinks of a way of acting (say, a way of playing a piano) in such a manner as to be surprised by the discovery that a given demonstratively presented (or explicitly described) way of thinking of that method of playing a piano is a way of thinking of the same method of playing" (ibid., p. 123).

Perhaps that is true, but what is the pianist doing when he thinks about a way of playing the piano?  As a pianist, I can offer my own introspected answer:  I might think about how to move my fingers and other parts of my body.  This would normally involve proprioception.  I am more likely to think about the look and sound of the keys, which is more auditory and visual memory.  I might also think about interpretative issues, like phrasing, tempo and so on, which can be more abstract, and not so much a matter of imagining myself producing or hearing the sounds.  A lot can be counted as ways of thinking about ways of playing the piano, and presumably many different ways of thinking can be involved. Perhaps a number of them are what Stanley would call "practical," but I'm not sure which.

Perhaps the most relevant sort of thinking about my performance (I think) is propriocepting a simulated performance.  This sort of activity might be generally categorized as employing a particular way of thinking and, yes, I think it is plausible that one can understand music in this way without being able to identify a given description or demonstration of what it is that one is thinking about.  However, I don't see how this helps Stanley's argument.

For one thing, it is not clear that Stanley wants to restrict "practical ways of thinking" to the proprioception of simulated motor activity.  If he did, then he would need an argument for why we should think that such ways of thinking are involved in actual performances, and not just simulations. Stanley does not think a conscious act of reflection must accompany the intelligent employment of one's motor skills. He balks at the idea that one must consciously avow the propositions which guide one's intelligent performance. But then, why associate a reflective capacity of proprioception with knowing how to act?

I must conclude that it is not clear how we are supposed to understand practical ways of thinking. Stanley's appeal to Peacocke is unfortunately not instructive. Peacocke writes: "What seems to be constitutive of these [action-based] ways of thinking is that when a thinker acts on an intention to perform an action thought of in one of these ways, he tries to act in a certain way (and does so in favourable circumstances)" (ibid., 124). I'm not sure what this means. Is the intention the object of thought, or is it the action which is thought of practically? Is Peacocke saying that one can act on intentions which are thought of in various ways, and one of those ways is action-based? In that case, we can act on intentions which are not thought of in an action-based way; in which case, being action-based has no consequences for whether or not an intention can be acted upon. What, then, does it mean to say that an intention is action-based?

Perhaps Peacocke means it the other way: The action is thought of in an action-based way. One may thus intend to perform an action which one thinks about in an action-based way, and the only way to carry out that intention is to carry out the action. However, Peacocke is not talking about what constitutes an intentional action. He is rather talking about what constitutes a practical way of thinking, and the way of thinking can be employed without carrying out the action, as when one reflects on one's abilities to perform. The pianist can think about ways of playing without actually playing. The actual carrying out of the intention to act cannot be constitutive of the way of thinking--in which case, Peacocke's view is self-contradictory.

Stanley's appeal to Heidegger is not more persuasive. Heidegger's claim is that when we use tools we engage more primordially with them. The suggestion is that our attitude becomes less intellectual or reflective and more active and procedural. The hammer, for example, is no longer an external object but a part of our own action. But again, does this show that knowing how to use a hammer requires the realization that one has this way of thinking about the hammer such that it is a good way to use the hammer? Or does knowing how to use the hammer just involve having the capacity to intelligently use the hammer, whether or not one has realized anything about one's way of using it?

In sum, the notion of practical ways of thinking is not clearly defined. It might refer only to the proprioception of motor simulations, but in that case it does not have a straightfoward connection to intelligent performance. Stanley's thesis seems to be that some intellectual reflection is necessary to turn one's practical thinking into practical knowledge, or know-how. Yet, he says no such reflection is necessary to employ one's know-how in practice. He says this because he knows that a regress argument is looming whenever such reflection is called upon. We might wonder, though, if a regress argument is still looming: If one must intellectually acknowledge one's practical way of thinking as a good way of thinking, mustn't one have a practical way of doing that? Mustn't one be skilled in the appraisal of that practical way of thinking? And so on, ad infinitum?

Practical and Theoretical

In a recent piece for The Stone, Jason Stanley rejects the common distinction between theoretical knowledge and practical skill.  His goal, apparently, is to liberate society from a false and oppressive elitism. He says: "The distinction between the practical and the theoretical is used to warehouse society into groups. It alienates and divides. It is fortunate, then, that it is nothing more than a fiction."  I'm not sure why that would be fortunate.  I rather think it would be fortunate if the warehousing turned out to be for the better.  If there was a problem, the fact that it was based on a fiction would not make me feel better.  In any case, Stanley does not draw our attention to any documented evidence of any alienation or division resulting from the folk conception of a practical/theoretical distinction.  I'm not convinced the folk conception is a problem at all.  But that's not what I want to write about right now.

Stanley's piece is written for a general audience, so we shouldn't expect the rigorous analysis or sensitive attention to detail that we would want from, say, his recent book, Know How (2011), which deals with the same topic but which is addressed to professional philosophers and advanced students.  We should only expect a very broad, accessible discussion of the topic, perhaps with some provocative examples or arguments.  Of course, we would not expect the main idea of the Stone article to directly contradict Stanley's more scholarly work.  And yet, that seems to be what we find.

I've finally gotten hold of an actual copy of the book and I've just finished reading Chapter 5, entitled "Knowledge How."  It is here that Stanley applies the linguistic analysis of Chapters 2-4 to provide a general account of what it means to know how to do something.  (By the way, now that I have the book, I see that I indicated the wrong page numbers in several of my recent posts.  I had been relying on chapters which Stanley has made available online, for people who have registered for his upcoming "Meisterkurs."  For some reason, those online texts have different page numbers.)  It is here that Stanley talks about the need for postulating practical ways of thinking.

Stanley differentiates between practical and demonstrative ways of thinking, such that the former are "action-based ways of thinking" (Stanley, p. 124) while the latter is the sort of thought that arises from "just staring" at objects (ibid., 125).  According to Stanley, this distinction is straightforward and obvious.  He says (italics in original)  "Anyone who accepts that cognitive states involve ways of thinking will have to accept practical ways of thinking" (ibid., p. 123); "As Peacocke and Heidegger point out, it is a straightfoward matter to construct Frege puzzles concerning something presented to one practically and that very same thing presented to one visually.  The only objections one could possibly have to practical ways of thinking of objects come from misunderstandings about the ontology of ways of thinking generally . . ." (ibid., 129).

As Stanley has it, one would have to be utterly confused to object to his appeal to the notion of practical ways of thinking.  Clearly our cognitive capacities are divided between the action-based and the "just staring" variety. The italics leave no room for doubt.

According to Stanley, there is a difference in cognitive capacities between being familiar and competent with descriptions of a way to do something and having practical knowledge of that way.  We can think and talk rationally about actions, reflecting on their ways, qualities and consequences, without having the practical knowledge required to apply those ways in action.  How is that not a distinction between theoretical and practical knowledge?  Yet, in The Stone, he argues against such a distinction.

In The Stone, Stanley introduces the theoretical/practical distinction as follows:

There is a natural temptation to view these activities [reflection and action] as requiring distinct capacities. When we reflect, we are guided by our knowledge of truths about the world. By contrast, when we act, we are guided by our knowledge of how to perform various actions. If these are distinct cognitive capacities, then knowing how to do something is not knowledge of a fact — that is, there is a distinction between practical and theoretical knowledge.
There are actually two distinctions here:  the theoretical/practical one (which he identifies with knowing that and knowing how, respectively) and the reflection/action one.  However, he seems to be mixing things up a bit.

What Stanley would presumably say is that we can reflect practically, and not just theoretically. Practical ways of thinking can be employed in reflection as well as action. For example, we might reflect proprioceptively on our motor skills, and thus reflect via a practical way of thinking. Thus, he might say, the difference between practical and demonstrative ways of thinking is not the same as the supposed difference between action and reflection. But then, why does he say that a sharp distinction between reflection and action would amount to a sharp distinction between practical and theoretical knowledge?  Indeed, his piece is called "The Practical and The Theoretical," and his entire thesis is that there is no difference in cognitive capacities between practical skill and theoretical knowledge.

In his book, Stanley argues that when we know how to do something, we have unique cognitive capacities which are not shared by people who only have detached, discursive or theoretical knowledge about that same activity or skill. Stanley explicitly argues for a sharp distinction between cognitive capacities: having practical skill involving an entity is not the same as having theoretical knowledge of that entity. Good plumbers have practical ways of thinking about plumbing; they utilize different cognitive capacities than people who are only capable of talking at length about plumbing. Yet, according to Stanley's Stone piece, this means that knowing how to do something is not knowledge of a fact--precisely the conclusion he wants to reject. Despite the noble attempt to demolish oppressive social barriers, Stanley's book supports the very distinction he claims is false.

Saturday, June 23, 2012

"What are you doing?" and other idioms

Interesting philosophical issues can arise when we investigate what it is we know or perceive about our actions while we are acting intentionally.  It should probably go without saying that, when we are acting intelligently, we know what we are doing, at least to some extent and in some sense of the verb, "to know."  I'm going to try to motivate the thesis that what we know is not obviously propositional in nature.

Consider the following question:

(1) What is she doing?
It goes without saying that, in many normal situations, a proper response to (1) is a description or identification of an action.  However, (1) has an idiomatic usage.  It can be used to communicate fear or anxiety about what somebody is doing.  It can communicate a desire for an explanation for an action.  Context usually makes it clear which usage is intended.

Now consider what happens when (1) is embedded in the following knowledge ascription:
(2) She knows what she is doing.
It's possible to use (2) to say of one person that she knows what another person is doing.  However, I want to consider the case in which we are stating of a person that she knows what she herself is doing.  In this case, we would interpret the ascription differently.  That is, we would not take it as a statement that a person who is doing something knows that they are doing it.  We would more normally take it as an ascription of another kind:  That the person has a good reason for doing what they are doing or, more, that their performance can be trusted.  This relates to the second, idiomatic meaning associated with (1), and not the literal meaning we might have expected.

A doctor is operating on a patient and an attending nurse cries out, "What are you doing?"  The doctor responds:  "I am about to make an incision here." The nurse says, "But that's the wrong spot."  The doctor responds, "I know what I'm doing."

The doctor knows what she is doing, which is not to say that she knows that she is about to make an incision in a particular spot.  It is to say that she is competent in what she is doing.  She is expressing confidence in her ability to perform properly, and not her knowledge that she is doing what she is doing.

The most natural reading of (2), when used to attribute a person's attitude towards their own person, is as I have outlined, or so I contend.  The logical conclusion is that natural readings of some knowlege ascriptions relate to idiomatic uses of their embedded questions.  Sentences like (2) indicate limits to the application of compositional semantics of embedded questions.  If my argument is correct, then we can conclude the following:  Not all knowledge ascriptions are properly analyzed by a compositional semantics of embedded questions.

I'm particularly curious about how this might have implications for ascriptions of knowledge how.  The well-known folk distinction between factual knowlege and practical knowledge is controversial.  Yet, the distinction is reflected in the idiomatic noun phrase, "know how."  It may well be that, thanks to this idiom, popular attributions of knowledge how resist a compositional semantic analysis.  On the other hand, it might be that the idiomatic noun phrase, "know how," is a natural result of the meaning of ordinary attributions of knowledge how.

Let's go back to the beginning for a moment.  Perhaps the natural reading of (2) is not the result of an idiomatic usage of questions like (1).  Perhaps the idiomatic use of (1) is a result of the natural reading of (2).  Thus, people might say "What are you doing?" to express concern because statements like "I know what I'm doing" express confidence in one's performance, and not the other way around.  Similarly, perhaps "know how" is a popular idiom expressing non-propositional competence because statements like "He knows how to ride a bike" express that sort of competence, and not knowledge of a natural answer to the question, "How could you ride a bike?"

Consider the following scenario:  Sam is told not to ride a bike, but disobeys the command.  Sam's parent is angry, and asks , "How could you ride a bike?"  The question is not asking for information about a way to ride a bicycle.  It is asking for a reason for one's having ridden a bicycle while also expressing condemnation of the act.  This idiomatic usage, however, does not show up in knowledge how ascriptions.  We do not use "I know how to ride a bike" to express that one has ridden for a morally proper reason.  However, we do use it to express that we have abilities required to ride a bicycle.  Yet, we would not use "How could you ride a bicycle?" to ask if somebody had the skill required to ride a bicycle.  The relationship between idioms, natural uses and embedded questions is neither simple nor obvious.

Is "I know how to ride a bicycle" naturally used to express knowledge of a w such that w is a way for one to ride a bicycle?  Or is it naturally used to express that one has the abilities required to ride a bicycle?  These questions are irrespective of whether or not such abilities count as a legitimate knowledge state and, if so, whether or not that state is propositional.  The point is that the use of know-how and other knowledge ascriptions do not clearly follow a compositional semantic analysis of emebedded questions, and it is not obvious what the natural, as opposed to the idiomatic, reading should be.

Musical Interlude: Chopin's Fantasie-Impromptu

My first video performance, and it's by no means flawless.  But it's the best I can offer for now.  A short improvisation followed by an imperfect rendition of Chopin's Fantasie-Impromptu:

Wednesday, June 20, 2012

No Gettier Cases for Know How

In Chapter 8 of Know How (2011), Jason Stanley discusses whether or not there can be Gettier cases for knowledge how, and whether or not this means knowledge how is not propositional knowledge.  He accepts the following proposition:

(P1) Gettier cases for know-how, if they exist, require that the subject intelligently and successfully F, where F ranges over actions.

(P1) is supposed to be self-evident, but this surely depends on how we understand "intelligently."  What counts as an intelligent action, in the sense that makes P1 self-evidently true?  Is it enough for the action to have been consciously decided upon?

We might be tempted to read P1 as only requiring that F be successfully performed with the conscious decision to F, which would entail the following proposition:

(P1*) Gettier cases for know-how, if they exist, require that the subject consciously decide to and successfully F, where F ranges over actions.

However, (P1*) does not seem self-evidently true.  One can consciously decide to F and still successfully F by accident, without having a sufficient set of beliefs about what it takes to F.  If there is not a sufficient set of beliefs guiding the action, then it would not seem to be a candidate for a Gettier case.

The only reason I can imagine why Stanley would think (P1) was plausible is because intelligent action involves being guided by rules, and Stanley takes this to imply a relation between a person and a proposition.  If one's successful performance is guided by rules for F-ing and if this entails a propositional relation (as Stanley believes), then we might have a candidate for a Gettier case.  The question would then be, does the propositional relation count as an expression of a justified true belief and, if so, is that justified true belief also a genuine case of knowledge?

Stanley accepts (P1) because he believes intelligent action is guided by propositional knowledge, and this means that it is more than just being successful and the result of a conscious decision.  In fact, Stanley seems to believe that if one's successful performance of F is guided by propositional knowledge, then that knowledge is knowledge how to F.  In other words, Stanley should accept the following proposition, retaining the sense of "intelligently" which makes P1 self-evident:

(P2) If one can intelligently and successfully F, then one knows how to F.

Yet, Stanley rejects (P2).  He does so in an attempt to undermine Ted Poston's (2009) argument that know how cannot have Gettier cases.  Stanley agrees with Poston that, if we accept P1 and P2, then there cannot be Gettier cases for know how.  Stanley wants to leave room open for such Gettier cases, and thus he rejects P2.  He explains why he thinks P2 is false by referring to a scenario described by Bengson, Moffett and Wright (2009; henceforth "BMW"):

Irina, who is a novice figure skater, decides to try a complex jump called the
Salchow. When one performs a Salchow, one takes off from the back inside edge
of one skate and lands on the back outside edge of the opposite skate after one or
more rotations in the air. Irina, however, is seriously mistaken about how to
perform a Salchow. She believes incorrectly that the way to perform a Salchow is
to take off from the front outside edge of one skate, jump in the air, spin, and land
on the front inside edge of the other skate. However, Irina has a severe
neurological abnormality that makes her act in ways that differ dramatically from
how she actually thinks she is acting. So, despite the fact that she is seriously
mistaken about how to perform a Salchow, whenever she actually attempts to do a
Salchow (in accordance with her misconceptions) the abnormality causes Irina to
unknowingly perform the correct sequence of moves, and so she ends up
successfully performing a Salchow.
 BMW posed this scenario to 138 people, and only 12 percent said that Irina knows how to do the Salchow.  BMW also asked about whether or not Irina has the ability to do the Salchow:  86 percent said she has the ability, but not the know how.  In response to these results, Stanley makes the following claim (Know How, 2011, p. 240):

In one sense of “intelligent”, Irina’s act of doing the Salchow is intelligent. It was
the result of a conscious decision. So in one sense of “intelligent”, Irina can intelligently
and successfully do the Salchow. However, if ordinary reactions about cases are granted
evidential weight, one must concede that Irina does not know how to do the Salchow.
Stanley thus concludes that P2 is false.  But notice that Stanley drew this conclusion by interpreting "intelligent" as "the result of a conscious decision." This would mean he is taking (P1) to actually mean (P1*).  But that can't be right.  (P1*) is not plausible and does not fit with Stanley's view of knowing how.

Let's consider Irina.  Why do BMW's subjects say she does not know how to do the Salchow?  I'll leave aside the possibility that the results are flawed.  Let's say they do reflect the common way of making know how ascriptions.  Why, if Irina can intentionally do the Salchow, do people say that she does not know how to do it?

Let's consider a different notion of intelligent action.  Ryle (The Concept of Mind, 1949, Chapter 2) presents the following notion of intelligent action: An action is intelligent, not by virtue of a conscious decision to perform, but by virtue of the way in which the performance is executed. If it is done with care, attentively, so that one learns as one goes and is able to anticipate and deal with novel situations and problems, then one's performance is intelligent. One must be guided by rules, but not passively and unthinkingly executing them. One must F mindfully. One cannot simply decide to F and hope for the best. That might be an intelligent decision (depending on the situation), but it does not confer intelligence to the subsequent performance.

Stanley's view of intelligent action is similar to Ryle's. The only difference is that Stanley takes the mindful guidance by rules in a successful performance to entail a knowledge relation between a person and a proposition.

Given the Rylean sense of intelligent performance (whether or not we take it to entail propositional knowledge), we must ask: Can Irina intelligently and successfully do the Salchow? She can do it successfully, there's no doubt about that. However, her ability to attend mindfully to her performance is obstructed by "a severe neurological abnormality" which gives her the wrong idea about what she is doing while she is performing. She cannot perform with the sort of care and learning that, according to Ryle, is the hallmark of intelligent action. This explains why so many subjects deny that she has the know how. In sum, BMW's results are consistent with P1 and P2, if we take "intelligently" in the Rylean sense, regardless of whether or not it entails propositional knowledge.

Stanley has not given a compelling argument against P2. On the contrary, his view of intelligent action seems to compel him to accept P2. Yet, by accepting P1 and P2, Stanley should conclude that there cannot be Gettier cases for know how.

Interestingly, Stanley might not be put off by accepting this conclusion. He does not think there has to be Gettier cases for knowledge how. In Know How (Chapter 8, pp. 240-244), he argues that many knowledge-wh relations are propositional in nature and resist Gettier cases. So he can try to explain Poston's result by observing that knowing how is just a case of knowing-wh, and that we should therefore not expect Gettier cases for it.

I don't think it's that easy. For one thing, Stanley has no explanation for why some cases of knowledge-wh are not open to Gettier cases. I have an explanation: Such cases of knowing-wh entail non-propositional know how. To take on one of Stanley's examples (p. 242), a tennis player knowing when to move a certain way does not thereby know a proposition, but rather how to follow a certain rule for playing tennis. That is know how, and that is why such cases of knowing when cannot be Gettiered.

Stanley merely claims that such knowledge-wh attributions are of propositional knowledge, and that this is uncontroversial. Maybe it is uncontroversial, but I question it. As a result, I can explain why such cases cannot be Gettiered.

(I'll also mention as an aside that the above argument is not committed to any particular position about Gettier cases, except that they are curious situations that can arise when we attribute propositional knowledge. I am not committed to Gettier cases posing a threat to the conceptualization of propositional knowledge as justified true belief. I think it is obvious, however, that Gettier cases only arise when dealing with ascriptions of propositional knowledge.)

Tuesday, June 19, 2012

Transparency at UVa

Helen Dragas has spoken in defense of her and the Board of Visitor's decision to force Terry Sullivan to resign from the position of President of the University of Virginia.  No reason was given for the move, and Sullivan had been a highly respected and appreciated leader.  Her resignation was a shock, but the most poignant demands were not that she had to be reinstated, but that the decision to remove her had to be explained.  Without transparency, only the most careless and vile motivations (e.g., corporate greed) can be supposed.

Dragas apologizes for this. She says, "our actions too readily lent themselves to perceptions of being opaque and not in keeping with the honored traditions of this University. For that reason, let me state clearly and unequivocally: you - our U.VA. family - deserved better from this Board, and we have heard your concerns loud and clear."

So you'd think she would try to close the gap and offer transparency, at least some. But she doesn't. Dragas does not offer an explanation. She claims that Sullivan was a good leader and that they are grateful for her service, but that they had to let her go. Dragas implicitly denies having to account for the Board's decision. In fact, she explicitly says that the board, and only the board, is in a position to know what is best: "the Board is the one entity that has a unique vantage point that enables us to oversee the big picture of those interactions, and how the leadership shapes the strategic trajectory of the University."

Her argument, in a nutshell: The Board of Visitors has privileged access to the right information, which allows them and them alone to properly evaluate the reasons for their decisions. The beloved UVa family should trust them, because their hearts and minds are well-intentioned and of the right stock. This can only be read as an argument against transparency.

Dragas denies the need to be transparent, and yet says the UVa "family" deserves better. I think they do.

Sunday, June 17, 2012

Opaque Contexts and Knowing How

Updated on July 1, 2012.  See bottom.

On page 229 of Know How (2011), Jason Stanley argues that ascriptions of knowing how create opaque contexts, which (he claims) implies that knowing how to do something is conceptual.  An opaque context is one in which co-extensive terms cannot be substituted without changing the truth-value of an expression.  Stanley is responding to the relatively familiar view that procedural knowledge or knowledge how is non-conceptual.  If it is non-conceptual, then why would it create opaque contexts?

Stanley quotes David Carr to prove that knowing how ascriptions create opaque contexts:

Suppose a famous dancer was to perform before an audience, an item from his
repertoire to which he has himself given the following title:   
(12) A performance of Improvisation No. 15   
To the astonishment of a member of his audience who just happens to be an
expert on communications, the movements of the dancer turn out to resemble an
accurate (movement perfect) semaphore version of Gray's 'Elegy', though the
dancer is quite unaware of this fact. We may describe what is seen by the
audience member as follows:   
(13) A semaphore recital of Gray's 'Elegy'

although we can describe the dancer as knowing how to bring about (12) we
cannot reasonably suppose that he also knows how to bring about (13). Even
though (12) and (13) are…but different characterizations of the same action, we
cannot safely switch these characterizations in knowing how contexts. So it
appears that sentences about knowing how, unlike those about ability, are truly
Stanley concludes:
A performance of Improvisation No. 15 is the same event as a semaphore recital of
Gray's 'Elegy'. But knowing how to do one does not entail knowing how to do the other.
Carr’s example shows that one may know how to ĭ without knowing how to Ȍ, even
though ĭ-ing and Ȍ-ing are the same actions. 

I find this wholly counter-intuitive.  Imagine the dancer, after performing, is confronted with the astonished member of the audience, who says, "I didn't know you knew how to perform a semaphore recital of Gray's 'Elegy.'"  The dancer might properly reply, "But I don't know how to do that!"  The audience member may then say, "But you just did!"  In which case, the dancer would rightly say, "A ha!  How funny!  I didn't realize it."

Stanley is, or at least should be, open to the idea that learning how to do something is acquiring a skill.  Yet, it would be absurd to say that, by virtue of this brief conversation, the dancer acquired the skill of performing a semaphore recital of Gray's 'Elegy.'  Yet, we can say that the dancer has learned that his dance is a semaphore recital of Gray's 'Elegy.'  He can now perform the dance under that description.  So should we say the dancer knows how to bring about (13) after the conversation, but not before?  Did the dancer acquire a new skill by learning that the same dance can be known under a different description?

If (12) and (13) are really coextensive (which is hard to imagine, but okay), then why not say the dancer knows how to bring about both?  The dancer might not know that he knows how to bring about both, which is why he needs to be told that (12) and (13) are coextensive before he can bring about the action as a response to a request to bring about (13).  But he still has the know how.

What is clear is that the dancer knows how to perform the action designated by (12) and (13), but only knows the action under one of the two descriptions.  This is a limitation in his declarative knowledge, not his know how.

We might suppose that the fact that (12) is an improvisation is important.  Maybe the implication is that the dancer, in improvising, does not have a clear conception of each of the movements, and so could not produce them on demand.  Maybe that is why we are supposed to think he does not know how to bring about (13).  The audience member might tell him that he had just brought about (13), and ask him to do it again, and the dancer might not be able to.  He might improvise on the theme designated by (12) and fail to bring about (13).  But in that case, the action types designated by (12) and (13) are not co-extensive, and so the argument for an opaque context collapses.

Either way, there is no opaque context here.  I don't see how one could argue for opaque contexts with respect to knowledge how, which suggests knowledge how is not propositional.

UPDATE:  I have a few more thoughts about why (12) and (13) are presumably not coextensive.  It is entirely plausible that they have different intentional properties.  Unlike (12), which entails an intention to improvise on a certain theme, (13) entails an intention to communicate something about Gray's 'Elegy.'  What defines (12) and (13) are not just the bodily movements carried out during the performance, but also causal-historical factors which define their intentional properties.  It is reasonable to assume that intentional properties supervene over or in some other way extend with respect to causal histories, in which case (12) and (13) are not coextensive.

Saturday, June 16, 2012

Stanley on Cognitive Neuroscience

I'm still working my way through the required reading for Jason Stanley's upcoming Meisterkurs at Berlin School of Mind and Brain.  I'm currently reading Chapter 7 of his Know How (2011).  At the moment, he's talking about how cognitive neuroscientists conceive of the distinction between declarative and procedural knowledge.  His argument is that the distinction has no implications for the question of whether knowing how is propositional in nature.

Stanley makes some very questionable moves here.  First, he discusses the distinction between declarative and procedural memory.  He is appealing to Gabrieli, 1998, p. 90:

nondeclarative or procedural kinds of memory encompass the acquisition, retention,and retrieval of knowledge expressed through experience-induced changes in performance. These kinds of memory are measured by indirect or implicit tests where no reference is made to that experience. Skill learning, repetition priming, and conditioning are classes of implicit tests that often reveal procedural memory processes dissociable from declarative memory.
The distinction is between memories which can be "consciously and intentionally recollected" (ibid., p. 89) and those which cannot, but which affect intentional behavior in other ways.  Stanley objects that we can take this account and still think of procedural knowledge as propositional.  He writes (p. 211):
Procedural knowledge is knowledge that is typically expressed through an increase in
skill. But knowledge of propositions could easily be expressed through an increase in
skill. My belief that I should catch a fly ball by positioning my body in a certain way
could become knowledge by practicing catching fly balls in that manner. The practice of
catching fly balls in that manner would give me proprioceptive evidence that my belief is
true. Unless one thought (absurdly) that proprioception is not a source of evidence, it is
obvious that propositional knowledge can be expressed through an increase in skill.
I want to give this argument careful attention, because it seems invalid.  More, it actually hurts Stanley's position.

One of Stanley's premises is that practice (e.g., the practice of catching fly balls in a particular way) can give one evidence in support of one's belief (e.g., that a particular method or rule is helpful when catching fly balls.)  I see nothing wrong with that premise.  So it does seem that practice can help turn belief into knowledge.  So, clearly, proprioception can be a source of evidence. But Stanley also has this premise:  that, given that proprioception is a source of evidence, it is obvious that propositional knowledge can be expressed through an increase in skill.

Not only is it not obvious that propositional knowledge can be expressed through an increase in skill; this premise is not consistent with the first premise, which states that propositional knowledge can be gained by observing that same increase in skill.  We can gain propositional knowledge by observing how we exercise a skill, as Stanley says, but that cannot mean that the exercising of the skill expresses the propositional knowledge it was supposed to justify.  If one has expressed their propositional knowledge that X by showing an improvement in their skill, then one would certainly not need to (and would not even be able to) use proprioception of that skill to move from mere belief that X to knowledge that X, because one would already have knowledge that X.  The fact that we can gain knowledge that X from observing our skilled performance implies that the skilled performance cannot be a manifestation of our knowledge that X.  Thus, either it is not a manifestation of knowledge at all, or it is knowledge of some other variety.

After this curious misstep, Stanley attempts to argue that, just as the two kinds of declarative knowledge (semantic and episodic) can be analyzed as different kinds of propositional knowledge, so we might expect procedural (nondeclarative) knowledge to be just another kind of propositional knowledge.  He suggests that they might be propositions which describe ways of doing things.

Now, obviously, we can have semantic and episodic memories about ways of doing things.  So why oppose that category of knowledge to declarative knowledge?  He mentions something about amnesiacs:

the procedural knowledge that an amnesiac can acquire is garden-variety
knowledge of a special class of propositions, those describing activities. For example, an
amnesiac who can acquire a pattern-analyzing skill can acquire knowledge of
propositions involving ways of doing that skill, but cannot acquire knowledge of many
other kinds of propositions, such as propositions about her past experiences. Even if
differences in how knowledge is implemented do correlate with kinds of content, the
contents of procedural knowledge are still propositions.

However, this is not an argument for the conclusion that procedural knowledge is propositional knowledge.  It is merely just stating that conclusion, and in a very counter-intuitive way.  For, again, why can't propositions about ways of doing something be semantically or episodically recalled?  And why should we think that what she has acquired is propositional knowledge?

Stanley also suggests that the way the declarative/procedural distinction is framed in AI might be applied here, so that we can suppose that what cognitive scientists call "procedural knowledge" might involve the sorts of representations that AI researchers call "declarative knowledge." Since Stanley thinks the AI distinction is not between two different types of knowledge, but between two different ways of implementing propositional knowledge, Stanley concludes that the cognitive neuroscientific distinction between declarative and procedural knowledge is not necessarily a distinction between propositional and non-propositional knowledge.  Yet, as I argued in my last post, Stanley has gotten the AI distinction wrong.

Still, perhaps Stanley could have a valid argument here.  The fact that cognitive neuroscientists distinguish between declarative and procedural knowledge on one level (the level of conscious access) does not exclude the possibility that these different kinds of knowledge might both have underlying propositional structures.  The AI distinction can be of help to Stanley here.  He can suppose that we do have declarative knowledge (in the AI sense) underlying what cognitive scientists call procedural knowledge.  It's just that we lack conscious access to the propositional contents.  In that case, however, the relation between a person and a proposition which, on Stanley's view, constitutes propositional knowledge must consist at a sub-personal level, through the base computational structure of the brain.  It is a relation between a person and that person's programming, and not a relation at the personal level of self, agency and intellect.  This is assuming, of course, that people are at base computational machines which implement programs the way AI declarativists think, and that is far from given.

At his most modest, Stanley only claims that we can interpret the cognitive neuroscience literature as being consistent with a view of procedural knowledge as a variety of propositional knowledge.  However, he is not always at his most modest.  For example, his conclusion on page 213 is a little shocking:
The obvious way to take the neuroscience literature is in the terms I have described. The
content of procedural knowledge is propositional, but involves different kinds of
propositions than stock cases of declarative literature.
Why is that obvious?  The cognitive neuroscience literature is continuous with the psychology literature, which for decades has drawn the declarative/procedural distinction in a very different way.  For example, Anderson (1976, p. 78) clearly identifies procedural knowledge with "knowledge about how to do something" and declarative knowledge with "knowledge of facts about the world."  Clearly the most obvious way of reading this distinction is not to take it as between different varieties of propositional knowledge, unless one were to think it obvious that all knowledge was necessarily propositional in some shape or form.

Stanley says (p. 213), "there is nothing in cognitive neuroscience that entails that procedural knowledge is not a species of genuine propositional knowledge."  I am not sure what Stanley means by "genuine propositional knowledge," nor am I convinced that this statement is a faithful representation of the field of cognitive neuroscience.  Perhaps Stanley can find a way to interpret the cognitive science literature as allowing for a propositional reading of procedural knowledge.  Perhaps there is enough support in the field for a declarativist view of knowledge implementation.  However, he has not shown that the literature is entirely amenable to such a view.  He has, on the contrary, engaged with an incredibly small portion of the literature.

Later in Chapter 7, Stanley discusses verbal reports of knowledge how.  He mentions a boxer who cannot verbalize his way of fighting a southpaw very well, but can vaguely identify it by performing and saying, "This is the way I fight against a southpaw" (Stanley, p. 219).  Yet, Stanley claims that reports such as this are not sufficient for declarative knowledge.  Being able to verbally declare that one's behavior is a way of doing something is not, in Stanley's view, a declarative competence.  Why not?  Apparently, because an indexical is given, and not a detailed description of the behavior.  But so what?

Clearly the linguistic centers of the brain must be involved.  There must be some linguistic competence there and it must be connected to whatever procedural knowledge is being implemented.  Yet, the linguistic competence is not identical to the boxing skill.  So clearly, there is declarative competence which can be distinguished from the procedural competence.  And clearly, being able to verbally identify a way of fighting against a southpaw is quite different from being able to implement that way of fighting against a southpaw.

It's odd that Stanley claims the boxer is not exhibiting declarative knowledge, when on the previous page, he observes that there is no neat way of characterizing declarative knowledge as cognitive scientists conceive of it:

Just as it is an open question how to characterize the notions of “implicit” and
“explicit” knowledge, it is equally an open question how to characterize the somewhat
technical notion of declarative knowledge as it occurs in cognitive neuroscience.

If the nature of declarative knowledge is still an open question in cognitive neuroscience, then why prejudicially assume that verbally identifying one's skills is not an exercise in declarative knowledge?  I don't see any argument for this in Stanley's tome.  He supposes that declarative knowledge might be knowledge that can be accurately and informatively described.  But isn't the boxer informing us when he identifies his skill verbally?  Isn't it an accurate statement?  Stanley (pp. 220-221) claims that what counts as informative and accurate is context-sensitive, but he has no doubt that the boxer's statement simply does not count, full stop.  Again, I don't see an argument supporting this counter-intuitive assertion.

I'm still working through Chapter 7 of Know How, and then I have Chapter 8 waiting, as well.  So, more to come.  But first, a little reflection on methodological assumptions.

On the one hand, there is the question about what various attributions of knowledge entail.  What do people mean when they attribute knowledge how to do something?  On the other hand, there are questions about the kinds of knowledge we actually have, and if all of them are propositional in one way or another.  Do they involve having conscious access to propositions?  Do they involve having organizational features which unconsciously utilize propositional knowledge states?  Do they imply an ability (or counterfactual ability) to verbalize or linguistically identify facts?  We might find that all kinds of knowledge involve some kind of relation between a person and a proposition, but that different kinds involve very different sorts of relations.  Yet, this will not mean that attributions of knowledge how to do something are actually attributions of knowledge that something is the case.  Not if we find that people attribute these different kinds of knowledge in independent ways and if they imply different relations between persons and propositions.

The fact that all human knowledge might be propositional in some base, computational way will not mean that knowing how is a species of knowing that.  And a linguistic analysis of the features of knowledge attributions is not the sort of study which should help us figure out how our brains actually work.  There is no reason I can imagine to suppose that the linguistic features of knowledge attributions would reflect the underlying computational scheme of our cognitive processing, or that we can infer from the grammar of the constructions to the operational structure of our brains.  It is much more likely that the meaning of these terms is related intimately to how they are used, and that they are used to distinguish between behaviorally distinct capacities.  This is what the folk distinction between knowing how and knowing that is about.  It follows something more or less like the psychologist's distinction between procedural and declarative knowledge, and not something like the AI distinction between different ways of implementing knowledge.

Stanley on the Procedural/Declarative Distinction

In Chapter 7 of Know How (2011), Jason Stanley discusses the origins of the distinction between procedural and declarative knowledge.  He argues that the distinction has never had any implications for the question of whether or not knowing how to do something is a kind of propositional knowledge.  In other words, he says the distinction has never implied that knowing how is different in kind from knowing that.  I think he's quite wrong and his argument misrepresents the literature.

Stanley appeals to a paper by Dienes & Perner (D&P) called "A theory of implicit and explicit knowledge" (1999), in which they claim that the distinction originated in AI and was later utilized in psychology.  On page 743, they claim that the distinction "concerned how best to implement knowledge."  They continue:

Should one represent the knowledge that all men are mortal as
a general declaration “for every individual it is true that if that
individual is human it is also mortal”? Whenever knowledge of a
human individual was introduced in the database this general
information would be consulted to infer by general inference rules
that that individual must also be mortal. The alternative would be
to have a specialised inference procedure: “Whenever an
individual is introduced that is human, represent that that individual
is mortal.”
These are two different procedures for producing the same output, which is a truth-evaluable representation of an individual as mortal.  Thus, Stanley concludes that the distinction between procedural and declarative knowledge is not a distinction between truth-evaluable and non-truth-evaluable knowledge.  It is not a distinction between propositional knowledge and non-propositional knowledge, but between different ways of programming a computer to produce the same propositional knowledge states.

Though he does not seem to realize it, Stanley is contradicting D&P's analysis, which claims that procedural knowledge is unique in that it only implicitly contains what is explicit in the declarative version.  They write in the next paragraph, which Stanley ignores:
The analysis also brings out the intuitive meaning of declarative
knowledge as knowledge that declares what is the case (e.g.,
Squire 1992, p. 204: memory whose content can be declared)
because it represents explicitly that something is a fact. Non-
declarative memory can be given precision in our analysis either
as the stronger form of knowledge that does not make predication
explicit or as a weaker form of knowledge that makes predication
explicit but leaves factuality implicit.
When D&P talk about different ways of implementing knowledge, they are talking about different ways AI programmers can implement their own knowledge in the programs.  They can build machines with procedural knowledge or they can build them with declarative knowledge, or both.  In some cases, these can be different ways of deriving the same propositional knowledge states.  Stanley's thesis, however, is that that is always and only what the distinction amounts to.  He ignores the very obvious fact that procedural and declarative knowledge can be different ways of producing intentional action.  The output of the system can be a non-propositional (i.e., not truth-evaluable) performance.

This is explicit in Winograd (1975), whom both D&P and Stanley appeal to.  On page 189, Winograd writes:
It is an obvious fact that many things we know are best seen
as procedures, and it is difficult to describe them in a purely
declarative way. If we want a robot to manipulate a simple
world (such as a table top of toy blocks), we do it most
naturally by describing its manipulations as programs. The
knowledge about building stacks is in the form of a program
to do it.

Winograd goes on to explain that the assumptions of the programmer are "built in" and not explicitly represented in or consulted by the system. He argues that this, the procedural kind of knowledge, has advantages "not only to obvious physical processes like moving blocks around, but equally well to deductive processes like playing games or proving geometry theorems."  So it is clear.  Procedural knowledge and declarative knowledge are different ways of producing behavior or output, but not necessarily the same propositional knowledge states.  The implication of the distinction is that there are two kinds of knowledge relating to two different ways of producing outputs.

Stanley seems to misunderstand this.  On page 208 of Know How, he writes:  "Clearly, Winograd is thinking of the distinction as one between ways knowledge is derived, not  kinds of knowledge states."  I would say that is clearly incorrect.

Winograd clearly and explicitly says that he is talking about different forms of knowledge.  For example, on page 190, Winograd writes:  "It is theoretically possible to express second-order knowledge in a declarative form, but it is extremely difficult to do so outside the context of a particular reasoning process."  And on the next page: "By putting knowledge in a primarily procedural form, we gain the ability to integrate the heuristic knowledge easily."

Winograd is distinguishing between systems which have knowledge built in through programs, and systems which have knowledge represented in a declarative way, as general statements which must be addressed in order for the knowledge to be implemented.  These are different ways of expressing or manifesting knowledge in a machine.  That is, they are different ways of implementing the knowledge we want the machine to have.  This is a distinction between how knowledge is represented in a system.  It's a difference in the kinds of states the system has.  And it's not just a matter of systems which derive propositional knowledge.  It's much more general.

The distinction, as it originated in AI (according to D&P and as demonstrated in Winograd), is this:  machines can have knowledge in the form of procedures without declarative content, and they can also have knowledge in the form of generalized, truth-evaluable declarations which they consult in their operations.  This does seem to have implications for a discussion of whether knowledge how to do something involves states or content which is not truth-evaluable.  Proceduralists believe that all human knowledge is at root procedural, and without propositional content.  Declarativists believe that all human knowledge is at root declarative, involving statements of fact.  If knowing how is purely procedural, it does not rely on machine states, code or instructions with truth-evaluable content.  However, Stanley erroneously concludes on page 208:  "As researchers in artificial intelligence use the distinction, it is irrelevant to the question of whether knowing how to do something is a state with a truth-evaluable content."  This goes against both D&P and Winograd, which exhausts the AI literature Stanley is consulting.

I will address Stanley's discussion of the procedural/declarative distinction as it pertains to psychology in another post.

Sunday, June 10, 2012

Counterfactual Statements of Fact

In "Are There Propositions?" (1929), Ryle interprets the word "proposition" as more or less synonymous with "statement." (See Ryle's Collected Papers Volume 2: Collected Essays 1929-1968, p. 39).  A logical result of this view is that propositional knowledge (aka "knowledge that") is knowledge related to statements of fact, linguistic or in other ways symbolic.  It would then follow that a person knows that x if and only if that person has some explicit recognition of the fact that x (via a representation of x), or the capacity to acknowledge or communicate that fact in some way.

We can understand capacities in terms of counterfactual abilities.  Thus, if a person knows that x, that person can, in some situations, explicitly recognize, acknowledge or communicate that fact in some way.  In this way, propositional knowledge is tied intrinsically to intellectual and discursive capacities.  It's no wonder, then, that Ryle identifies propositional knowledge with "the jobs of didactic discourse" in the ninth chapter of The Concept of Mind (1949) or that, in the second chapter of that book (as well as in his 1946 paper, "Knowing How and Knowing That"), he talks about propositional knowledge in terms of avowing, reciting or considering propositions.  Ryle's discussion of propositional knowledge seems consistent.

Ryle contrasts propositional knowledge with knowing how:  knowing how to do something is having a complex of dispositions and capacities to perform intelligently, where intelligent performance is marked by creativity, active learning, improvisation, and care.  This is much more than what is implied by mere propositional competence.  When a person acts intelligently, they are guided by rules--their behavior is not unthinking, but it is not necessarily intellectual, either.  It does not necessarily involve the consideration or even the acknowledgment of statements of fact.  For Ryle, being able to intelligently follow a rule is categorically different from being able to acknowledge or state a truth.  Ryle offers numerous regress arguments to prove that the former cannot be reduced to the latter.  (For a formal analysis of Ryle's regress strategy and related topics, see here.)

In his book Know How (2011), Jason Stanley critiques Ryle's regress arguments, but he agrees with the point that intelligent performance does not require the avowal or acknowledgment of any rules.  He does not, however, believe that knowing that is a matter of abilities to do with the acknowledgment or avowal of facts.  He thus challenges Ryle's view of knowing that.

I've argued in other posts at length about how I disagree with Stanley's interpretation of Ryle.  Now I want to offer a Rylean criticism of Stanley's view of propositional knowledge.  I'll focus on two examples which Stanley uses.

In one example, Stanley claims that propositional knowledge is exhibited through the opening of a door by the turn of a handle.  When a person successfully opens a door, we can attribute propositional knowledge:  They know that that is a way to open the door.  In the second example, a person might not be able to recall the sequence of numbers in a passcode, but they might still be able to enter the correct numbers into the keypad.  Thus, we can say they know that those are the right numbers, even though they cannot verbally pronounce them or otherwise identify them.  Stanley's conclusion is that knowing that something is the case is not limited to knowledge which can be discursively shared.  He says propositional knowledge is not always a matter of didactic discourse.

A Rylean would respond by supposing that the capacity to acknowledge or communicate the relevant facts is present, even if the subject cannot acknowledge or communicate them in the present instance.

In the case of opening the door, we are attributing common knowledge.  Everybody knows that that is a way to open a door.  It's something that anybody could identify and discursively share.  Any normal person could communicate, in some way or another, that that is a way to open a door.  We don't require a demonstration of their skill before we attribute this propositional knowledge.  In fact, I don't think a demonstration of competence at opening doors would even help.

Consider that cats occasionally figure out how to open doors by jumping and turning the handles.  We do not so easily say that the cat knows that that is a way to open the door, do we?  We much more easily attribute knowledge how to open the door.  When we say a person knows that that is a way to open a door, this is not because they have successfully opened a door, but because we think they're in a position to know that they did it in a particular way.  Cat's might have the know how, but they're not so obviously in a position to acknowledge or communicate the facts.

The passcode example is a little different.  In this case, we might assume that the subject has momentarily forgotten the passcode.  They once learned it and now they have to exert themselves a bit to remember it.  Surely they can remember it somehow.  They might, for example, simulate the pressing of the keys on the keypad, or they might try to remember something else which might trigger the memory of the numbers.  Either way, we attribute propositional knowledge because the subject can recall it with intellectual effort, even if they cannot recall it off hand.

There are presumably some cases where we attribute propositional knowledge even when we don't suppose the knowledge was explicitly learned as such.  Perhaps, if a person can use mental simulation or reflection to identify some fact about a procedure they know how to perform, then we are justified in attributing propositional knowledge of that fact.  The person, again, is in a position to know, and can counterfactually recognize or acknowledge the fact, even if they have never done so.  It's also worth noting that it is probably quite difficult, perhaps impossible, in many cases to decide if a person's ability to identify a fact about a known procedure is a matter of recalling a previously learned fact or if it is a case of identifying it for the first time.

Considering all of this, it is not so hard to understand the temptation to say that procedural knowledge is implicitly propositional--that if you know a procedure, you have some implicit propositional knowledge of the way to do it.  In Ryle's view, this is an error.  The fact that we can, in some cases, acquire explicit propositional knowledge from reflecting on or simulating procedures does not mean that our procedural knowledge is reducible to propositional knowledge.  It does not mean our procedural knowledge is just an implicit form of the explicit propositional knowledge.  It does not mean that all procedural knowledge is propositional knowledge.  All it means is that some procedural knowledge entails implicit or latent propositional knowledge.

The Rylean view might be summarized as follows:  A person knows that x if and only if they have the counterfactual ability to correctly acknowledge a statement of the fact that x.  This ability is neither necessary nor sufficient for attributions of knowledge how.

Perhaps Ryle is wrong.  Perhaps not.

Friday, June 8, 2012

Stanley on embedded questions and related topics

I've just finished the first chapter of Stanley's Know How (2011).  He ends by summarizing Ryle as follows: "According to Ryle, one knows how to do something if and only if one bears an intellectual relationship to an action type."  This is wrong in two ways.  First, Ryle does not suppose all knowing how is intellectual.  That's precisely why he frames his version of knowing how against intellectualism.  Not all intelligence is a matter of intellect.  Second, Ryle explicitly denies that knowing how can be defined in terms of particular action types.

In any case, Stanley makes a very strong argument on pages 48 to 51.  Ryle (1949) claims that the fact that we do not say "believes how to do something" is evidence for a strong type distinction between knowing how and knowing that.  Stanley observes quite persuasively that the verb "believe" does not take embedded questions, and so this is not a peculiarity of constructions involving "know" plus "how."  We do not "believe where something is" or "believe why something happens," he notes.  So, if his version of intellectualism is correct, the fact that we don't believe how to do things is only evidence of the fact that attributions of knowing how contain embedded questions.  (He does not attempt to explain why the verb "believe" does not take embedded questions.)  Furthermore, if such attributions do contain embedded questions, Stanley argues, then knowing how to do something is presumably knowing the answer to a question.  This directly contradicts Ryle's claim that knowing how to do something is not knowing the answer to a question.

This analysis also helps Stanley explain why it is we intuitively believe that there are gradations of knowing how.  What we are grading are the relative values or contents of different possible answers to embedded questions.  Not all correct answers are equal in value or content. Thus, we may grade knowing how without supposing that the knowledge itself has gradations.  This directly challenges Ryle's observation that, unlike knowing that, knowing how comes in degrees.

Stephen Hetherington (in "Good Knowledge, Bad Knowledge," 2001, as I recall) has similarly challenged Ryle's argument from gradation, arguing instead that both knowing how and knowing that come in degrees.  We can say both that a person knows how to do something very well and that a person knows very well that something is true.  But Stanley will presumably respond that such judgments are comparisons between different instances of knowledge, and not evidence of gradations in knowledge itself.  I'm not sure what kind of evidence could be found to support Stanley over Hetherington, or vice versa.  It comes down to how we conceptualize knowledge and what it would mean for knowledge to be graded.  Either way, the point against Ryle stands.

Stanley's argument on these pages is compelling.  However, as I noted a long time ago, the fact that attributions of knowing how contain embedded questions might well mean that the attributed knowledge is knowledge of an answer (or set of answers) to a question, but it does not force the conclusion that the attributed knowledge is propositional, because not all answers to questions are propositions.  Some questions can be answered with an intelligent performance, and not the recitation of a truth.  Thus, we might make room for a mention none reading of some knowledge attributions.

Furthermore, there is another linguistic difference between knowing how and knowing that which Ryle suggests, but which Stanley does not account for.  I have discussed it before (see the link in the preceding paragraph).  It is that there is an asymmetry between "How do you know how to X?" and "How do you know that Y?"  The first question asks for an account of a skill or competence.  The second asks for a justification for a belief.  If we accept the answer to the first question and the person does know how to X, and if our acceptance of the answer is justified, we do not thereby know how to X.  Yet, if we accept the answer to the second question and the person does know that Y, and our acceptance is justified, then we know that Y.  If knowing how were simply a matter of knowing that, this asymmetry should not appear.  This is a point Ryle suggested, but did not fully develop, in his (1949), when he says that we do not ask for reasons for one's know how, but only for one's knowledge that.

Ryle's Lewis Carroll Argument

In his 1946 paper, "Knowing How and Knowing That," Ryle makes brief mention of Lewis Carroll's 1895 paper, "What the Tortoise Said to Achilles."  (I discussed Carroll's argument and its relation to Ryle not so long ago.)  Jason Stanley addresses the problem in his recent Know How (2011).  He quotes Ryle's argument, which is not quite identical to Carroll's.  Ryle writes:

A pupil fails to follow an argument. He understands the premises and he
understands the conclusion. But he fails to see that the conclusion follows from
the premises. The teacher thinks him rather dull but tries to help. So he tells him
that there is an ulterior proposition which he has not considered, namely, that if
these premises are true, the conclusion is true. The pupil understands this and
dutifully recites it alongside the premises, and still fails to see that the conclusion
follows from the premises even when accompanied by the assertion that these
premises entail this conclusion. So a second hypothetical proposition is added to
his store; namely, that the conclusion is true if the premises are true as well as the
first hypothetical proposition that if the premises are true the conclusion is true.
And still the pupil fails to see. And so on for ever. He accepts rules in theory but
this does not force him to apply them in practice. He considers reasons, but he
fails to reason. (This is Lewis Carroll’s puzzle in ‘What the Tortoise said to
Achilles’. I have met no successful attempt to solve it).

Ryle concludes:
…knowing [a rule of inference] is not a case of knowing an extra fact or truth; it
is knowing how to move from acknowledging some facts to acknowledging
others. Knowing a rule of inference is not possessing a bit of extra information
but being able to perform an intelligent operation. Knowing a rule is knowing
how. It is realized in performances which conform to the rule, not in theoretical
citations of it.
 According to Stanley, Ryle's interest here is in whether or not epistemic justification amounts to having only "the right propositional states" (Stanley, p. 41).  I don't think that's right.  For one thing, Ryle never talks about anything called "propositional states."  He is wary of the notion of literal mental states and he was highly critical of the contemporary notion of propositions.  Ryle is interested in whether or not our accepting logical arguments can be reduced to our accepting a set of sentences.  Furthermore, Ryle is not focusing on whether or not one might be justified or not in one's reasoning, but rather on whether or not one is compelled to accept a conclusion given certain set of sentences.  So the focus is not so much on what constitutes epistemic justification, but on what compells us to accept certain logical arguments.

Ryle attempts to explain the absurdity of the example by claiming that reasoning is not reducible to knowledge of truths.  The Rylean moral is that reasoning is not exhaustively modeled by a sequence of sentences representing the truths which are reasoned, and that if you try to model it that way, you end up with an infinite regress.  Propositional knowledge alone does not force us to accept or draw logical conclusions.

Stanley makes the following objection.  He supposes that Ryle is working under the assumption that propositional knowledge is behaviorally inert.  That is why such knowledge supposedly does not compel the student to accept that the conclusion follows from the premises.  Accepting that the conclusion follows from the premises would be a behavior, and Ryle's point is supposedly that the student's propositional knowledge does not facilitate any behaviors.

This is a mistake.  Ryle does not take propositional knowledge, or any other kind of knowledge, to be behaviorally inert.  In fact, Ryle's hypothetical student does produce behaviors based on his acceptance of the relevant propositions.  He "dutifully recites" them and "considers" them.  For Ryle, consideration, reflection and contemplation are forms of behavior, and their subject matter is propositional.  Why should we think Ryle supposes that propositional knowledge was behaviorally inert?  The problem confronting Ryle's hypothetical student is not that his propositional knowledge has no behavioral inertia, but that it does not have one very specific behavioral consequence:  the acknowledgment that the conclusion follows from the premises.

Ryle distinguishes between theory and practice, but not such that only the latter is behaviorally operant.  Ryle's claim is that, if a person only knows something in theory, they do not know how to apply it in practice.  Ryle takes theoretical knowledge to be behaviorally limited, not inert.  Theoretical knowledge involves certain sorts of dispositions and competences, and not others.  Though Ryle's vocabulary is a bit outdated, his view is not so foreign.  Today the distinction is more commonly drawn as between language and motor abilities and between higher and lower cognitive functions.  There is good reason to appreciate such distinctions and to acknowledge their neurological bases, and I don't think Stanley wants to doubt them.  He only fails to see how Ryle's view is a precursor to this contemporary way of thinking.

Stanley contrasts Ryle's view with Boghossian, who argues that understanding the logical constants may just be a matter of being disposed to and having the capacity to apply them in certain ways.  But Ryle would certainly find some sympathy with Boghossian's assessment, though he would insist that the sort of competence Boghossian is talking about cannot be reduced to the justified true belief in propositions.

Stanley's objection at this point is likely thus:  There is no reason to think of these distinctions as implying the non-reducablity of knowing how to knowing that.  The crux is that Stanley takes the notion of propositional knowledge to be wide enough to encompass all kinds of knowing.  Either these other competences and abilities are knowing that, or they are not knowledge, period.  The fact that formal logic does not exhaustively model human reasoning is probably obvious to Stanley, and not in need of lengthy philosophical argument.  Yet, in Ryle's defense, his Lewis Carroll argument wasn't all that lengthy.  Moreover, the fact that we do not need this and similar arguments today might be in part due to the fact that Ryle so forcefully made them.

Thursday, June 7, 2012

Jason Stanley "Know How" Meisterkurs

In early July, Jason Stanley will be giving a three-day Meisterkurs (in English) at Humboldt University's Berlin School of Mind and Brain.  With his blessing and the permission of the organizers, I have registered and plan on attending the event.  It's going to be hard, because I have no sympathy at all for Stanley's interpretation of Ryle, which is likely going to take up a large portion of the course.  Discussion has been encouraged (even in German, which means I might not understand everything that's said), but I don't want to bother him or the other students with countless objections.  Still, half the required reading is either on or by Ryle, so I'm going to have a hard time biting my tongue.  I'll try to measure my comments carefully and keep them short and to the point.  I would love to go head to head with Stanley on Ryle, and perhaps I will have a suitable opportunity over the course of the three days, but I'm not going with that kind of agenda.  I would much rather spend the time trying to figure out where Stanley and I agree (or disagree, though I'm rather keen on finding some common ground) on more general issues, like how to understand propositional knowledge and the relationships between linguistics, epistemology, cognitive science and the philosophy of mind.

Now that I've registered for the course, I have access to the readings online.  They include three chapters from Stanley's recent book, Know How, which circumstances have prevented me from purchasing.  I've thus been able to find yet more problems with Stanley's interpretation of Ryle.  I'll describe one for now.

Stanley quotes the following passage from Ryle's 1946 paper, "Knowing How and Knowing That":

If a deed, to be intelligent, has to be guided by the consideration of a regulative
proposition, the gap between that consideration and the practical application of
the regulation has to be bridged by some go-between process which cannot by the
pre-supposed definition itself be an exercise of intelligence and cannot, by
definition, be the resultant deed. This go-between application process has
somehow to marry observance of a contemplated maxim with the enforcement of
behavior. So it has to unite in itself the allegedly incompatible properties of being
kith to theory and kin to practice, else it could not be the applying of the one in
the other. For, unlike theory, it must be able to influence action, and unlike
impulses, it must be amenable to regulative propositions. Consistency requires,
therefore, that this schizophrenic broker must again be subdivided into one bit
which contemplates but does not execute, one which executes but does not
contemplate and a third which reconciles these irreconcilables. And so on forever.

This is a clear case of hypothetical reasoning.  Ryle begins with "if a deed . . .," because (as the context makes clear) he does not agree with the stipulation.  Specifically, he is trying to draw logical consequences from the intellectualist view of intelligent action, which he explicitly rejects.  Thus, later in the paragraph, when he refers to the "allegedly incompatible properties of being kith to theory and kin to practice," we should remember that it is the intellectualist who alleges this incompatibility.  Ryle does not make this allegation, as careful reading of his work should make clear.  It is the intellectualist, in Ryle's view, who supposes that theory and action are so at odds.  This is precisely the view he is trying to overturn.

 Indeed, on the same page, Ryle writes the following:

It also helps to upset the assumed type-difference between thinking and doing, since
only subjects belonging to the same type can share predicates. But thinking and doing do
share lots of predicates, such as ‘clever’, ‘stupid’, ‘careful’, ‘strenuous’, ‘attentive’, etc.

Recall that, in Ryle's understanding, it is the intellectualist who wrongly assumes that theory is where intelligence is found, and that action is only intelligent by virtue of its association with theoretical acts. Ryle, in contrast, denies this supposition.  Thus, when Ryle says that, "unlike theory, it must be able to influence action," Ryle is not presenting his own view of the relationship between theory and action.  Ryle rather says that reflection and contemplation are varieties of intelligent action.

Thus, as I said, the passage which Stanley quotes shows Ryle trying to draw the intellectualist view to its logical conclusion.  It is hypothetical reasoning, an attempt to draw problematic consequences from the intellectualist view of thought and action, and not a presentation of Ryle's own views of thought and action.  Stanley is therefore egregiously mistaken when he says on page 40, after quoting this passage:  "Ryle assumes here that possession of propositional knowledge is behaviorally inert. In other words, Ryle assumes that the possession of propositional knowledge does not entail the possession of dispositional states."  Ryle makes no such assumption.  Ryle is not here presenting his view of propositional knowledge.

I find myself disagreeing with Stanley's interpretation of Ryle at every turn.  This is not just a difference in emphasis.  It's not confined to one or two isolated passages.  Stanley consistently and profoundly gets Ryle wrong, so much that he accuses Ryle of believing the very opposite of what Ryle explicitly claims.

As an attempt at objective adjudication, one might ask the following question:  whose interpretation of Ryle is the most consistent and charitable?  On my account, Ryle is consistent and, if not always right, at least always thought-provoking.  On Stanley's account, Ryle is constantly contradicting himself and often times absurdly flying in the face of common sense.  Charity clearly falls on my side.

I have more reading to do to prepare for the course.  Unfortunately, this June's a terrible month for me, so I'm a bit frustrated with the timing.  Still, I'll try to organize my thoughts as much as possible, and I plan on writing more blog posts to help.  So expect more thoughts on Stanley, Know How and related topics in the coming weeks.