Dissertation Archives

Donald Davidson

Donald Davidson, the philosopher of mind and language at Berkeley, died yesterday. This makes the second giant of philosophy -- especially of my particular philosophical experience -- to die in the last 12 months. Rawls passed away earlier this year while I was in France. As a major figure of the canon of political philosophy, he was on the front page of the Times and many papers. Donald Davidson, though, was a philosopher's philosopher at the heart of the tumultuous "linguistic turn" of 20th century philosophy. The big advances in this century were not in ethics or political thinking, they were in understanding the role of talking and thinking themselves.

It's up to you to decide what the big advances of the last century were, not me, but I'll continue giving you my humble opinion.

When Williams James at Harvard and John Dewey at Columbia were pushing the pragmatist agenda in the early 1900s, they were moving in a current that advanced thinking about the mind. The first formal study of psychology was solidifying its sometimes strange results and the agenda for philosophy of mind became urgent. James was into all kinds of strange things, like séances and mindreading, but the future held a series of a crises around the issue of mind and body: how can the physical sciences explain everything if the mind is not a mere physical thing? And if the mind is a physical thing, how can the mind be as free and powerful as we thought it should be? A later Harvard professor named B. F. Skinner claimed to have captured the mind on a page of formulas, precise and empirical relationships between inputs and outputs.

Meanwhile, a European line of inquiry from the late 19th century sought to formalize with the precision of mathematics that most essential discipline: reason itself. If we can organize and precisely study the structure of mathematics, why can't we do the same for arguments and reasoning itself? We should know the truth of an argument from its pure and unsullied structure. Formal logic's father was Frege and its famous pioneers for anglophone philosophy were Whitehead and Russell. While philosophy was typically conducted in ordinary sentences (as hard as they sometimes were to understand) the dream of people inspired by Russell and Frege was to create a philosophy that was perfectly formal: If P and P-->Q, then Q. Stuff a computer could check.

Indeed, computers were being built to operate in precisely this fashion, and a result of this line of research was Turing's test for when a computer had successfully implemented a mind. In fact the underlying assumption is more than just a test; it is an argument that a computer can implement a mind as surely as it implements any complex program code. (The Church-Turing thesis. You can also look at Chapter 1 and 3 of my dissertation.)

A computer is a machine that operates observing the complex formal instructions of its programs, languages that were designed following from the developments of Frege's "calculus" for logic. And a computer running such a language can implement a mind. A computer can be a mind. (Maybe every computer is always a mind, where some minds--like ours--are smarter than others.)

I'm being rather informal here, but the main idea is that language and mind were important top-level currents in 20th Century philosophy through war.

When Davidson arrived at Harvard, studying with the great W.V.O. Quine (who died in 2000), he was on to be influenced by the important arguments of the 1950s against logical positivism and empiricism. Essentially, the move began against the view that language could be made like logic. The logical positivists thought that we could avoid nonsense sentences and stick to simple stuff, like "This snow is white." Adding some logic to build up such little sentences into big theories (like "All snow is white."), we could make logical language the building block of careful science. Quine and Davidson pushed a campaign against this view.

It's folly to think you can ever be careful or precise enough to achieve the purity of mathematical precision. Complex sentences can't simply be broken down into simple ones like "this snow is white". Nor can we distinguish sharply between the observed facts and the "logical" language. Sometimes the logical language is just as much up for debate or disagreement as the observations -- neither is pure or precise. In fact, Davidson seemed to push this further. He argued that it is silly to say that there is some "mental picture" on the one hand and there is some actual reality on the other hand. The very idea that science is about getting our mental picture to be faithful to reality just sets us up to fail.

That's why Richard Rorty at Stanford, the dean of American Pragmatism, loves to group Davidson with Derrida and Wittgenstein. Very different figures! It's a provocative grouping that usually annoys philosophers in philosophy departments. It sure annoys the guys in the Stanford department. But he's talking about something that any undergraduate can appreciate: all these guys think language and the things we say are separate from the "world". They don't have to match and there's no way to know if they do. Language has its own internal rules, roughly, and figuring out what people mean depends on what you think they mean or what they could possibly mean. It doesn't depend on the specific chemical structure of some tiny particle. "Reality" is obscure and pretty much irrelevant to these guys. We all decide what's true or false using what we think is good evidence.

Davidson's work was in language and mind. I ran into this language stuff as an undergraduate when I studied with Akeel Bilgrami at Columbia, himself a student of Davidson's and a solid Davidsonian. He got me thinking Davidson might be right about meaning -- it's not the apple in the bowl that tells us what "apple" means, it's what I'm thinking when I say apple. It's what I mean by "apple". You see perhaps how this technical argument in the philosophy of language connects with Rorty's point about truth. You can judge my sentence "All apples are red" without knowing what I mean, even if you already see a green apple in the bowl.

I read the mind stuff at Stanford, with another Davidson student, Michael Bratman. The gist of Davidson's idea was to say that our mental life pretty much consists of our attitudes towards sentences. Well, it's not proprietary Davidson stuff. But this idea is the one wanted to work with to explain psychological questions. Why did Amol walk to the fridge? Because he believed a sandwich was in the fridge and desired to eat it. Two discrete attitudes, strict rules for how they interact, and bang! An action results. We can apply the rules of logic to understand the function of the mind, just as if it were a computer. Reasons are the causes of action just as the cue ball is the cause of another ball's motion.

I carried Davidson's picture (above) around in my wallet for about 4 years. So I figured he deserved some comment with his passing.

Graduation Pics

I graduated last week. Here are a select few pics.

With my advisor Mark Crimmins

With Etchemendy, Stanford Provost

Happy Birthday to Me

Yesterday was my birthday, and I published the acid-free final candidate of my dissertation. This is the one that should get signatures and go on the shelf.

Acid-free Final

Now for a Professional Achievement

I've just had the delightful news that a paper of mine has been accepted to a conference this summer in Sydney! The Australasian Society for Cognitive Science's conference this summer at UNSW. I'll give my folk psychology paper, the one based on chapter 3 of my dissertation. It's 13 - 17 July 2003. Tell me if you'll be in Sydney that week.

Making Corrections

The defense was a great experience. It made me reconsider my decision to totally abandon the academic path. One concrete result: I will certainly submit a paper based on Chapter 3 of my dissertation to some journals. I will revise it a bit now, show it to some folks in the area, and then send it in to Mind & Language. Depending on how that process goes, I may be drawn to submit my file for some jobs this fall. Maybe, but not likely. I don't think I want to be an assistant professor anywhere. And after working in consulting for a while, that interest may be even lower.

The keys to a defense are:
* scheduling the thing
* preparing some handouts - one to outline your presentation, and some to deliver as backups on difficult questions
* controlling the agenda with your opening remarks
* hitting back when the cranky guys start making noises

I got a surprisingly large turnout for that time of day, 9am on a Friday. And I really think the students bought my general line on many questions. That is, the professorial attack didn't stick. I even made some progress in convincing those guys themselves of my overall position.

The discussion was so enjoyable that I'd love to have more discussions like it in the future. People suggested a dozen new directions to take from the work I had done, and pointed out all kinds of interesting results from it. They made it feel significant, worthwhile, and a progressive program. The kind of stuff to build a career on.

Hence the sudden renewed optimism. But I don't know if it's the career for me from other perspectives. The problem was never that I wasn't good at it, or that I didn't like the activity -- it's just that the conditions of employment are so bad. Security, flexibility, teaching burden. It's a rough life on campuses.

At the moment, I'm responding to a list of corrections and concerns that people raised in the defense. Not much work, but it feels an awful lot like another round of revisions. Luckily, this time there is a finite end to all this. June 4, the filing date for the acid-free printed draft of this thing complete with signatures. Shouldn't be long now.

Oregon and Beyond

After those serious looking images of me accomplishing important things, I present you with a gallery of my little trip to Oregon, Portland, the Cascades, etc.

Oregon 2003 Gallery

The Defense Day

This is me after my dissertation defense on April 4th. In the background is Stanford's iconic Hoover Tower. I'm looking very serious for your benefit only. I'm actually very pleased: Things went well! At the motel in Oregon a few days later, I forgot to register as "Dr. Sarva". But I did learn to appreciate the pleasures of California sparkling wines made by the methode champenoise.

I'll stick up some pictures of the Oregon trip in a bit.

Defended!

done!

The Hardest Part of a Doctorate

I now discover the hardest part of this damn PhD. People sort of mentioned it, but it is damned difficult to schedule a time for your defense. 5 people, 1 room, 2-3 hours. Oh it's terrible!

It's been 2 weeks that the date has been bouncing around, and I'm trying to time buying my plane ticket. It's very frustrating. So bad, in fact, that I can't even concentrate on preparing for the bloody thing.

Milestones of the Last Five Years

The road to the Ph.D.

November 1997 - Application

April 1998 - Acceptance

September 1998 - Arrived

Winter 1999/2000 - Evolutionary Psychology Course
I took a course with Ken Taylor where we investigated a range of issues connected to evolutionary psychology, part of which was their use of modularity or "massive modularity". Looked at critiques of the massive modularity hypothesis.

Spring 2000 - Folk Psychology
Did a reading group with Peter Godfrey-Smith on folk psychology's debate between simulationism and theory-theory. Developed the idea that this debate turned on distinguishing between intentional modules and mechanism modules.

May 2000 - Coursework Completed
I completed all major components of the coursework requirement (I had one piece left till that Fall, but they let me go ahead anyway). That was the 2 year mark in the program.

Summer 2000 - Dissertation Development Seminar
To kick of the dissertation process, my class year took a seminar with two professors, including my eventual supervisor Mark Crimmins, where we would spend the summer searching for a dissertation topic. It didn't play out like that: the first day, they said "what have you got on your mind?" I gave an answer: "the modularity of mind". We were sent to do research, think about the topic, start writing a 4-5 page dissertation proposal.

I started reading some introductory-type and popular books on the subject to learn the literature. I started copying dozens of papers and buying lots of books. I was developing a bibliography and a sense for what the controversial issues were. I "presented" the proposal to the seminar twice, and each time developed it further. What I ended up with that August is by and large what I have worked from. The chapter topics have evolved.

Fall 2000-Spring 2001 - Third Year
During this year I had to do some TAing. I also worked to develop the core ideas of the dissertation. So I did a lot of reading. I also wrote one chapter, which has become Chapter 2 - Modularity and Nativism in Chomsky. It's a historical/context chapter which discusses a range of arguments to establish the background ideas.

Spring 2001 - The Chapter Roadmap
When PGS got back from a semester away, I met with him to discuss some ideas I'd been working on. I handed him a sheet of paper with about 5 or 6 paper ideas on it. 4 of those have become chapters in my dissertation. Of course, they changed a lot! But they have changed quite a bit.

Summer 2001 - Hiatus
Non-philosophy events took over my agenda at this point. Done with my TAing duties, I was finished with my need to be on campus. Late this summer, I actually packed up and headed for New York.

Fall 2001 - Slow Movement
Another few months where progress was pretty slow. I was reading, thinking, but maybe only 2 days a week. An idea that Mark Crimmins gave me at some stage was to look at a paper by Davies and Stone that became a crucial motivator for the next chapter I wrote.

April 2002 - Fourth Year Talk
We are meant to present a non-introductory chapter to the department. I wrote and presented my Chapter 3 - Folk Psychology and Intentional Modules. The chapter circulated to my committee members for a few weeks before this presentation, but I got very sketchy comments in general.

I wrote up the talk, rehearsed it a few times, and gave it to the department. It was tough! People were reasonable critical, didn't really understand a lot of what I was saying, didn't appreciate it, etc. So it was tough -- few people in the department are really conversant with the issues I was talking about.

April 2002 - Departmental Orals
The next day, we scheduled my Oral Exam. It is not supposed to come after, but it made sense given the logistics. Typically, I think it comes a few months beforehand. Since people would be listening to me present my chapter talk anyway, it made sense to do it afterward.

Biggest surprise was that people in the Oral pressed literally the same issues they had pressed in questions to my talk. It makes sense that people press what grabbed them, but it is surprising that there isn't a sort of "review" for quality. They just sit down and debate. It was rough; I didn't really feel I convinced anyone. But I passed, so I guess that is how such things go.

Summer/Fall 2002 - Writing the Chapters
Then I packed up for Paris, sat down and budgeted 6-weeks per chapter. This was an insane schedule -- Ch3 had taken 3 months. I came to Paris with the goal of writing 2 more major substantive chapters in only 3 months. This ran a bit over into September. Then I was "done" with my dissertation. Not really - I had to write the "introductory" chapter. After 3 months of working long hours 6 days a week, I took a month off.

I sent the drafts off to everyone. Didn't get any feedback on most things.

Winter 2002/2003 - Finalizing
It took about 3 weeks to write the introductory chapter in December. Throughout the Fall/Winter, I was preparing for the job process--academic and non-academic jobs. This had been going well but was a time investment.

By January, I had written the substance of all the chapters - 1 (introductory), 2 (historical/context), 3 (fourth year talk), 4 & 5 (written in summer).

Then I waited for feedback. In February, I had a series of discussions with my adviser to make revisions on the stuff. In early March, I sent out a new set of revisions. I sent out another new set in late/mid-March.

April 2003 - Defense
The date is now set for the defense. It was hard to finalize a date. Mainly scheduling, it seems like! We'll see how it goes. I am led to think that it is a "sure thing". Anyway, I have to present for 30 minutes and take questions for 2 hours. Should be hot and stressful. I wonder what I should eat beforehand?

I think I was underprepared for my 4th year talk. This time, I'm going to specifically prepare briefs to use in response to various lines of questioning. For example, I know the main concerns of the core committee. I should prepare written/diagrammed answers to them. I will also meet with various people in the days leading up to the defense to get their early reactions.

Spring 2003 - Filing the final dissertation
I'll have to make some more changes, I guess. Then, I'll file and have the document published. It'll go in the library and over UMI. You'll be able to buy it from dissertation.com or wherever.

June 2003 - Graduation
I guess this is when I get the robes.

The Concept of Modularity in Cognitive Science

You've been wondering. Now you can know:

The Concept of Modularity in Cognitive Science

Pushback

I was hoping that this last go-round would have nailed-shut the revisions open pre-defense, but it's seeming like folks are not totally comfortable. Too long, is one complaint. And another is that I haven't adequately addressed some key criticisms. It feels like my last round of changes didn't do much to address the worries there.

So, there is the possibility of pushing off the defense date a few weeks. This would be disappointing, so I hope it doesn't happen. Overall, though, I think the main objective to wrap things for June is still in place. It just means I may have to do more work to get there.

Writing PDFs

PDF995 is a really great, free PDF writer. I used it to encode my bibliography. I'll put the rest of the dissertation online pretty soon too. (In fact, it's already online but in a secret place. Can you find it? I don't really want to have Word docs of it available too widely; so when I put them into PDF, I'll share them as well.)

Dissertative Moods

In the taxonomy of PhD milestones, I'm going to take the risk of calling The Defense the most significant. It is the much-storied Defense that wins Super Bowls. Defense that is the best offense. Defensiveness, the most praiseworthy of sociable moods.

My dissertation defense, the "University Oral Exam", is coming up. My aiming for late-March right now, depending on the schedules of my committee members.

The last big achievement I was trumpeting was when I "finished" my dissertation. By that, of course, I meant that I had finished a draft of my dissertation, the first comprehensive draft of the main argumentative chapters. Since then I wrote a little introduction, took criticisms on the main sections, and thought about how it all fits together. I also made a test effort at the academic job market (which was moderately successful, assessed on its own terms).

Now last week, I received 3 days of commentary from my adviser, which I will incoporate in taking my dissertation final. From the day I make it final, at least 28 days must elapse before my examiners have to offer the Oral Exam. (When I'm out West, I'm going to drive out to AZ and check out the Cactus Leagues with an old philosophy chum.)

On that date, they could say "Cool, you're all set." More likely, they will offer a short list of vague directives on ways to improve what I have (or make it better cohere with their model of the world).

I'll make those changes, have my adviser approve them, and then I will have dispensed every academic requirement left.

I'll print up half a dozen copies of the thing, get them signed by the committee, and then get a bound copy or three made. And on June 15, I'll wear the hood and take some kind of oath and get the heck on with things.

The End of the Dissertation

I announced with great pleasure on September 15th that I had finished my dissertation. If not for the loose ends! I had to write an "introduction", which annoyed me like a dripping faucet for months. I polished it off Friday afternoon. It really is awful, but I don't much care about it anymore. It also turns out to be about 25 pages long. Glad to close that "chapter".

Shot in the Dark

It is partly vanity, "all is", but it is also partly genuine interest.

I sent off ten application packets for academic jobs this week, of which the plum assignments are tenure-track junior posts in mind or simply open at Rutgers and NYU. Some enviable positions at UCLA or CSULA, but those are not for me. The bottom bottom line is, when I say "interest" I mean curiosity. I'm confident I will receive zero interviews.

But it is worth mentioning that today I received confirmation that Ken Taylor's letter of recommendation was in. That in addition to one from Peter Godfrey-Smith last week and a teaching letter from John Perry. That leaves, of course, my advisor Mark Crimmins. When that comes through we send it all off. Who knows? Maybe they wrote brilliant things about me...in which case, I could get interviews after all.

Though I haven't even planned to attend the Eastern APA this winter.

Writing sample? I cut out some sections from my chapter on folk psychology and tacit knowledge. Some day, I'm sure you'll read it excerpted in the New Yorker. Ha!

Now I'm Done

Now I'm Done
September 10th, 2002 - I finished the last word of the last chapter, and now my dissertation is done. Of course, things to add, change, revise. But for the most part, we're at the top of the mountain. No sweat left. Now onto better things:

- Paris's teaming film life. So many films, every night, everwhere. Saw "Wittgenstein" (awful) and "Les Vacances de Mr. Hulot" (by Jacques Tati, not bad).

- Reading. Catch up on all those english books, and catch up on all that french grammar. Preparing for the Diplome de la Langue Francaise.

- The fall shows at the museums.

- Perhaps some sniffing for the coming job hunt.

Please send your congratulatory gifts to:
Amol Sarva
Philosophy Dept
Stanford, CA 94305-2155

Thousands in, inches left

Thousands in, inches left
I'm thirty pages into the final chapter now. It's very tough going. I've hit a wall where I am completely bored with the work under way. It's making me feel a bit nauseous in fact. I can hardly bear to look at it. Hopefully I can refreshed and return to it again.

The Meta-Structure of My Dissertation

The Meta-Structure of My Dissertation
I have had a breakthrough, 3 days ago now, with the structure of the final chapter. With that, I now can outline the overall dissertation. Of course, there is still the ugly work of writing this last chapter, but see below.

The Actual Structure
Title: The Concept of Modularity in Cognitive Science

Chapter 1. Modularity and Nativism in Chomsky
A look at the four major conceptual shifts that underpinned the shift from behaviorism, the psychological doctrine of B.F. Skinner or theory of mind of Gilbert Ryle, to cogntivism, the doctrine of modern scientific psychology. The key new ideas: the mind is a real object of study (cognitivism) not just a "way of talking" about pure physical events (behaviorism); the mind operates in informational, representational states that interact in systematic ways (representationalism, computationalism); many cognitive capacities have strong innate bases (nativism); the cognitive capacities are organized into interacting, distinct parts (modularism). This chapter traces the historical arc of this shift of thinking at a high level by reviewing the major arguments given by Chomsky and others, particularly for modularity and nativism.

Chapter 1.5 What is Modularity
This chapter articulates a concept of modularity for contemporary cognitive science. Since Chomsky and Fodor popularized this model of mind culminating in the early 1980s, the idea has been picked up and diversified. It has lead to misunderstanding and many conflicting usages. I review the specific features attributed to modularity, rejecting some and identifying a few as the core of the concept. By my conception: modularity requires informational isolation (a notion related to Pylyshyn's "informational encapsulation"), but does not require domain-specificity, neural localization, systematic breakdown, innateness, adaptedness, shallow inputs, peripheral position (i.e. Fodor's "only perception is modular"), or many other features. This is a radical simplification of the concept and tendentious insofar as it rejects domain-specificity as playing a constitutive role. But I think it improves the debate by focusing on what it could mean to be modular, and leaving contingent features aside.

Chapter 2. Tacit Knowledge in Folk Psychology
A major recent debate over folk psychology or mind-reading tracks a serious confusion about the nature of modules. One traditional distinction is between modular mechanisms and modules of knowledge. Mechanisms are cognitive capacities that perform certain functions, in the way a physical tool or device operates on a certain type of object. Knowledge on the other hand is something known but modular in that it is a distinct and isolated database. I argue that this distinction is what is driving the debate between the theory-theory and the simulation theory. Both sides are trying to distinguish their position from the other, but an empirical standstill persists. I claim this is because the distinction between mechanism and knowledge is faulty: a module described as a mechanism is just an account of the implementation of a knowledge module. Collapsing the distinction solves a key feature of the folk psychology debate, by concluding that mechanisms are just bodies of "tacit knowledge".

Chapter 3. How Modularity and Nativism Connect
Nearly every nativist theory of psychology is also modularist, with few though prominent exceptions. Why should this be? I consider the positions, argue that there is no intrinsic or logical connection between the positions. I argue that the two appear together for methodological reasons: the principal lines of argument that evidence nativism are in fact evidence for modularity, and vice versa. This is a delicate argument but very interesting, and explains a striking phenomenon in contemporary cognitive science.

Chapter 4. Three Types of Domain Specificity
A crucial concept in a variety of accounts of modules is domain-specificity, especially among developmental theorists and evolutionary psychologists. But while researchers in psychology use the term comfortably, it is difficult to isolate a single clear account. I argue that the literature entertains not one but three very different accounts of domain-specificity. I consider the advantages and disadvantages of these various accounts. I support one among these.

The Meta-Structure
The concept of modularity in cognitive science
1. History of the concept of modularity in the contemporary phase of cognitive science
- historical chapter
1.5 A proposal for what the concept should be, and a dismissal of some incorrect construals
- my positive proposal, plus a review and organization of the views already out there
2. A traditional distinction between two types of modules is mistaken; "mechanism" and "knowledge" are just two levels of description for a system as a crisis in folk psychology illustrates
- theoretically radical criticism about an empirical debate
3. Modules need not be nativism conceptually, but there is a good reason that theorists usually take up both
- novel observation of a link in some discussions, and a theoretical argument about its theoretical basis
4. Domain specificity is a confused concept; it can't be "the essence of modularity"
- inquiry into a central concept, clarification of the confused conceptual options

The great gears of Philosophy recommence their turning

The great gears of Philosophy recommence their turning
After some weeks of incubation, pen hit paper in the last few days. Starting out the next chapter. What will it be about? you gasp in excitement.

Domain-specificity -- a key concept for modularity theories and particular evolutionary psychology's brand of modularism.

Everybody's using it nowadays, and it seems that they do so without problems. Some things are clearly domains: language, folk pscyhology, mathematics, vision, audition, and then also perhaps cheater-detection, religion, natural laws like biology and so on. But what is domain-specificity?

Roughly, it is the idea that certain cognitive mechanisms are specialized to work on certain functions. You have a language system, a social reasoning system, a mathematics system, vision system, etc. You can find people with isolated impairments: no language comprehension, no language production, no vision (no vision for fruits...no vision for motion...other specific), no math, etc.

If you look for someone to tell you how preicsely domain-specificity is defined, however, you have problems. There are a limited number of mostly breezy efforts to give you the definition.

Some themes:

1. Logical vs. de facto character -- some things are domain-specific by accident of their position. They don't get access to the right problems or info or inputs because of where they are plugged into the overall system of processors. Other times some subject matters are simply domain-specific a fortiori. The subject matter or system can only be domain-specific because of its logically isolated nature.

2. Processors vs. subject matters -- There are cognitive capacities or devices, and there are the problems or topics that they concern. So think of hammers and nails. Hammers can be domain-specific -- facts about the hammer make it "specialized". Or you can define the domain specificity on the basis of what stuff it applies to -- only certain classes of nails. So if the nails are a subjet of all possible nails, that's proof that something is a domain-specific hammer. Alternatively the account can be based on facts about the hammer only. After all, you can try hammering normal nails with a pin-head hammer (even if it doesn't work well). The domain-specificity claim for it is based on its pin-head, not a real restriction on what stuff it can actually hit.

3. Vs. Adaptive Definitions -- sometimes theorists ignore both elements. They just give an account based on the "adaptive problem" to be solved. I think this is confused; it has to collapse into one of the elements of point 2 above.

4. Flexibility -- in general, it's a continuum. Not a simple "either its domain-specific or domain-general".

Some other stuff too. But that's the direction I'm heading. The goal is to sharpen it up and perhaps get rid of some of the criteria people are applying. Hopefully, a result would be to restrict the class of possible ways to be domain-specific. Wish me luck. The future of the human race depends on it!

Mud

Mud
It's hard to get going again now that the second-to-last chapter is done. And there's this atmosphere of gloom besides: what's the point? will the parts fit together? who cares anyway? even when I'm done where will I get a job? etc etc. It is, in fact, post-partum depression! Well, who knows. But the vagueness and irrationality of the worries rings of the classic 'blues' or 'writer's block' unhappiness.

I've been there before, so let's just keep this note as an historical artifact. One has ups and downs. See for example the note from July 11.

Measuring Proress in Spilt Ink

Measuring Proress in Spilt Ink
Here's how I've come so far: Progress Chart - August 2, 2002.

Look at the bottom of your screen to click on the charts. But you'll get the raw numbers first. As of this moment, I have 40,000 words or 160 pages of this thing in the bag. Once upon a time, I had planned to churn out only 125 and call it an education.

Progress and Next Steps

Progress and Next Steps
Last Thursday I finished Chapter 3, "How Modularity and Innateness Connect", where I claim that the arguments that imply each innateness and modularity also lead investigators to conclude for the other. So arguments for innateness, such as Poverty of the Stimulus, in fact imply that there is a module down there too. The massive chapter is 70 pages long or so, partly because it enumerates several arguments for each nativism and modularity.

I have now taken essentially 6 days off. 4 days in Berlin. 2 days thinking about post-dissertation jobs and stuff, putting in some applications. Now it's August 1 and I ought to get cracking again.

Next step: domain-specificity. This one may not be as long, but that's what I always say. Don't know what I'll say yet.

Dissertative Moods

Dissertative Moods
It's not all roses you know. Sometimes it feels like you're getting nowhere. Had a couple of strong days just now, so let's hope we keep ahead à tout.

Steps...Steps

Steps...Steps
Well, the direction for Chapter 3 has been plotted and it's not what you were expecting. Here's a rough outline:

1. Innateness and Modularity frequently turn up together. Some evidence that this is the case from various cites. (Fodor, Chomsky, Spelke, Khalidi, Cosmides et al., Keil, Pinker, Gopnik et al., Leslie, etc.)

2. What they are. What is innateness, roughly. (Chomsky) What is modularity, roughly. (Ch2 concept) Intrinsic independence of the two concepts; examples of how you can have one without the other in each direction.

[2.5. Intrinsic links. Consider reasons for thinking they are indeed intrinsically linked. Reject. It will be empirical.]

3. Innateness --> Modularity
3.1. Arguments for Innateness (general, of a capacity)
3.1.1. Poverty of the Stimulus (Content)
3.1.2. Developmental Fixity / Growth of Fixed Capacities, a.k.a. Poverty of the Stimulus (Properties)
3.1.3. Impossibility of Learning
3.2. Implications for Modularity
3.2.1. If Nativisim is warranted
3.2.2. Direct warrant for modularity

4. Modularity --> Innateness (similar structure to above)
4.1. Arguments for Modularity (general, of a capacity)
4.1.1. Dissociations, Neuropsychological locality
4.1.2. Informational encapsulation
4.1.3. Developmental program/schedule differences and independence
4.1.4. Uniqueness of the domain-specialization (won't work)
4.1.5. Adaptive history (Cosmides style)
4.2. Implications for Nativism
4.2.1. If Modularity is warranted
4.2.1.1. Poverty of the Stimulus for Modularity (Botterill/Carruthers, Khalidi)
4.2.1.2. Modularity is complex inner structure, which is a version of "developmental rigidity" arg for nativism
4.2.2. Direct warrant for nativism
4.2.2.1. adaptive history clearly --> nativism
4.2.2.2. dissociations could be plastic/non-innate (though non-regeneration --> innateness of structure/specialization?)
4.2.2.3. informational encapsulation suggests physical/biological/innate limits on scope of "reasoning"
4.2.2.4. developmental programs imply innateness directly
4.2.2.5. note about soft modules (chess skill) - penetrable, even though "modular"

Results

Results
Started writing yesterday. A few things are clear: a) the chapter 1 draft needs to be rewritten and the modularity section in there actually needs to be drafted, b) chapter 2 should be rewritten to focus more on the concept of tacit knowledge, and less on the debate in folk psychology, c) chapter 3 coming up could go a number of different directions and I'm still debating: mainly, responding to an interesting paper that sets up the discourse, or trying to survey a wide literature of cognitive psychology and neurpsychology work, d) I'll have to simple push forward, the area is vast, e) I probably do have the right books and materials (maybe it'd be nice to take another look at Fodor's "Present Controversy".

Report on the Last Two Weeks

Report on the Last Two Weeks
Since hitting the ground in Paris:
* Cracked open the _Psychology of Language_ text
* Elman et al. _Rethinking Innateness_ -- Ch1, Ch 7 arguments
* Fodor's review of Elman et al
* Fodor and Pylyshyn's 1988 "Connectionism and Cog Sci"
* Khalidi's 2001 and 2002 papers on nativism/domain-specificity and nativism/triggering
* Fodor's review of Cowie, Cowie's response
* Eric Lennenberg's 1964 paper on language innateness
* Hirschfeld and Gelman's "Introduction" to their domain-specificity book
* Sketched plans for thesis chapter 3 and chapter 4 (innateness and domain-specificity chapters)
* Reviewed Fodor 2000's domain-specificity stuff
* Davies "Does 5% Make a Difference" paper on Evans/Peacocke/Wright tacit knowledge stuff
* Sketched further ideas on the mechanism/content distinction in cog sci -- they're at different levels, and so orthogonal characterizations
* Not really needing the IJN/CNRS, since Pompidou and Mitterand are such excellent places to work (may use the library there if it comes to it)
* Will participate in the summer Mind and Cognition workshop at Sorbonne

Institut Jean Nicod, Centre Nationale de la Recherche Scientifique

Institut Jean Nicod, Centre Nationale de Recherche Scientifique
Visited with Pierre Jacob at the IJN, CNRS today. A wonderful fellow. Clearly, Nicod is a great place to be if you are working on philosophy of mind and psychology in Paris. Not sure how much I'll really benefit (resources-wise) from being affiliated with them over the summer, but apparently I'll be able to get access to some libraries and such. The Institut itself doesn't really have tons of rooms or library space to hang out in.

Attended Millikan's talk there as well, and this alerts me to the importance of looking back into Dretske's theory of representation. As Pierre Jacob pointed out also, perhaps I should look at Tyler Burge's discussions of partial concept possession to shed light on the issues of tacit knowledge.

Dissertation Update

Dissertation Update
I'm disappointed to announce that shareholders should expect two further weeks of stalled work. We are in full-bore moving mode now, then we'll be travelling, then we'll be getting settled in Paris. So it is difficult to imagine getting much done between now and then.

In the last two weeks, the plan has been to collect research materials, in anticipation of a semi-isolation from resources during the summer. To that end, I believe I have collected substantial materials.

I will carry one book with me and try to make progress on it: Elman et al.'s Rethinking Innateness. We'll see if that provides the right prod to my present thinking on nativism.

Areas of focus:
* nativism and psycholinguistics
* domain-specificity and evolutionary psychology
* tacit knowlege and folk psychology
* shaping a "modularity in dissaray" thesis chapter

Concerns:
* the tacit knowlege theme - lots of things may count as tacit knowlege (including simulating mechanisms). If that's so, then the way to say it may be: "tacit knowlege is a promiscuous concept". Leave it to the reader/audience to adjudicate the result, though my leaning is to say "fine, lots of stuff counts...and therefore lots of stuff is mental".
* domain-specificity - there isn't actually a good debate on the precise delineation of the concept. Here I may have to set out a view myself. Should also contact David Sloan Wilson as I haven't gotten his view in a while.
* connection of nativism to domain-specificity may be easier than connection to modularity itself. But that could be enough to yield an interesting thesis about the inter-relation of component parts of the modularity thesis. One problem is that this feels bound up right away with the domain-specificity chapter ("what is d-s?") in a way that presumes that a solution is offered. Only if you know what d-s is can you say whether d-s is bound up with nativism, right? Not sure. Hopefully they can stand separately.

Oh yes, and we have now launched a French Studies blog for thoughts on those matters.

Dissertation Update

Dissertation Update
In play:
* Elman et al. _Rethinking Innateness_
* Barkow, Cosmides and Tooby _The Adapted Mind_
* Still waiting on Jeffrey Samet to send along his stuff on historical conceptions of nativism.

Dissertation Update

Dissertation Update
A massive treasure-trove of interesting papers in _Cognition_ were discovered today. I guess Cognition must be the leading cognitive psychology journal. It's hard to know these things if people don't tell you, but this thing is monthly and it's constantly full of all the big names.

Many interesting debates, one particularly interesting involving Paul Bloom about the domain-specific facts required for word learning. Several other discussion of domain-specificity. Not so much of innateness in the last two years of papers that I've looked at so far.

One especially valuable find was Strevens's paper from Feb 2000 about several types of psychological essentialism. The real valuable bit is that its outline is a perfect "PGS paper" outline, i.e. "Here is a concept in science. Here is what it could mean. Here is the difference between these options. Done."

The outline:
1. Introduction
2. Varieties of essentialism
2.1. Pure essentialism
2.2. Statistical essentialism
2.3. Internal essentialism
2.4. The minimal hypothesis
3. How the essentialist hypothesis explains the data
3.1. Projection
3.2. Categorization
3.2.1. The discovery experiment
3.2.2. The transformation experiments
3.3. Comparison of projection and categorization
3.4. Other explanations of the K-patterns
3.4.1. Gelman's projection experiments
3.4.2. Keil's discovery experiment
3.4.3. Keil's transformation experiments
4. The minimal hypothesis
5. Against psychological essentialism
5.1. Against statistical essentialism
5.2. Against internal essentialism
5.2.1. Children's explicit views on the causal role of insides
5.2.2. Insides versus lineage in the discovery experiments
5.2.3. The timing of the discovery and transformation results
5.3. Against pure essentialism

Bold new plan! Finish thesis

Bold new plan!
Finish thesis by September. Chapter on Domain-Specificity and chapter on Innateness both due to be drafted by then. Plus a chapter on What Is Modularity.

Today's results:
Rewrote a sketch of the entire "FPs threat of collapse" paper while the ideas were still fresh in mind. Will return to this later to draft a publication article (if possible) and to further refine Chapter 2. Wrote about 10 pages just sitting there for a few hours.

Due next:
Will go for 5 mile run when I get home. Now training to hit 10 miles by mid-Summer.

Last Friday: finished the talk,

Last Friday: finished the talk, gave the talk, fielded some questions.
Last Saturday: got the laptop stolen, lost the talk.
Yesterday: passed the orals.
Now on to the revised thesis timeline: finish a full draft by September. That means at least two more chapters by the end of the summer. That's double-time.

Finished writing the talk on

Finished writing the talk on Sunday. It's half the length of the paper. Now I must also write up a sheet to use as a hand-out. Talked to Bloser and Eden over the weekend to get some feedback. Talking to Crimmins later today.

Writing up the talk now.

Writing up the talk now. Wrote a few pages. Need to figure out what to handout. I guess I'll write the talk first then figure out that bit. Should aim to have that all done by Sunday.

Today: tidy up the draft

Today: tidy up the draft a little. Then send off for some comments.

9,900 (includes cutting 1,000 of

9,900 (includes cutting 1,000 of notes etc that were lingering around).

I now declare this to be the first draft.

9,600 or so. We are

9,600 or so.

We are in the end game now of the threat of collapse section. I have set out Heal and D&S's views. Now I am in the midst of setting out my response.

One thing that doesn't look like it will fit is an elaboration of how this can connect to modularity-proper. How it can function to help with domain-specificity and so on. Dunno yet. This has been long so far!

7,590

7,590

Really need to finish this

Really need to finish this thing today if I'm going to get it read by next week. I've actually got to give the talk next Thursday! Soon!

7,359 (as of last night)

7,359 (as of last night)

5,990

5,990

5,804 (rewrote the 'trivialization of

5,804 (rewrote the 'trivialization of simulation' section)

5,328

5,328

Additional papers considered: * Heal

Additional papers considered:
* Heal 1996 (Mind and Language) - Simulation Cognitive Pentrability
* Nichols, Stich 1998 - Cognitive Penetrability, Restricted Rationality, and Off Line Simulation
* Perner et al. 1997 - Predicting Others by Simulation or Theory? A Method to Decide

Oh yes, another one of

Oh yes, another one of those distractions: S+B article is now dead meat. I need to invoice them for the kill fee. I suppose I should tweak it (more examples, more detail...longer) and shop it back to Mel and/or other publications.

End of Thursday: 3966. Another

End of Thursday: 3966. Another rough day. But we soldier on, as always. Today's goal is to enter the heart of the debate. The preliminaries in setting out FP, TT, ST are through. Now to handle the bases of distinction, the framework of computationalism and the varieties of collapse.

Wednesday close: 3400. Distractions! Meeting

Wednesday close: 3400. Distractions! Meeting with Alumni Relations, call with Renee, other disruptions.

By brute count, it's at

By brute count, it's at 3000 words now. So the net add is 1200 for today. But alot of it is "notes" that will be cut. Then again, I cut some other stuff. I'll just tally the net-brute-count daily. Should be indicative enough.

Results! * 1500-words yesterday -

Results!
* 1500-words yesterday - first three sections: Folk Psychology, Thy-Thy (partial), Simulation Thy (opening)
* 300-words today (so far) - added some sections to Thy-Thy, Folk Psychology
* Block, "The Mind as the Software of the Brain"
* Stich and Nichols, "Folk Psychology: Simulation of Tacit Theory" in Davies and Stone eds. 1995
* Davies and Stone, "Introduction" in Davies and Stone eds. 1995
* Nichols, Stich, Leslie and Klein, "Varieties of Off-line Simulation" in Carruthers and Smith eds. 1996
* Josef Perner, "Arguments for a simulation-theory mix..." in Carruthers and Smith eds. 1996
* Alison Gopnik, "Theories and modules..." in Carruthers and Smith eds. 1996
* Downloaded a bunch of Leslie's papers in .pdf from his lab site.
* Received: Kim Sterelny's The Representational Theory of Mind

Today! The best intentions were

Today! The best intentions were to build on the 250-some words I drafted yesterday. However, instead I read:
* Simon's chapter 1 (introduction) and
* Pylyshyn's chapter 2 (computational foundations of cognition) in Posner ed. (1991), "The Foundations of Cognitive Science".

It's a massive collection but pretty interesting.

One key insight from Pylyshyn is that the problem of "what is computation" is general and old, particularly analog vs. digital "implementations" of a function. Because a system is analog does not mean it doesn't implement a functional architecture that is equivalent to a Turing Machine (both extensionally equivalent and also algorithmically equivalent or "isomorphic"). Take this with the promiscuity of "computation", and every physical system implements some kind of automaton -- but the interesting ones will be Multiple State Automata rather than simply input-output mappings via Finite State Automata.

Another interesting point is that computation is broadly the set of physical systems that implement some kind of formal system which would include both "classical computation" (or Turing/von Neumann machines) and also connectionist networks (as well as other types of machines).

If the folk psychological "threat

If the folk psychological "threat of collapse" goes through, this should be taken as a significant result for the distinction between intentional and computational modules.

Working on advancing the issues

Working on advancing the issues for the completion of Chapter 1, Section 4 (modularity's history and general currents) as well as setting out for Chapter 2 (folk psychology and the nature of modules). Heal has posted a copy of her 1994 paper (hard to find). I will begin by looking back through the stuff in Carruthers and Smith 1996 (including Segal's modularity paper and the Davies/Stone review of the debates).

Interestingly, Segal takes up the thy-thy view in his discussion of the nature of the folk psychology module. The interesting question for me is how the simulationist view is any different from the hypothesized thy-thy view. So if the simulationist facility (the rider box that lets you repurpose practical reasoning into folk psychological reasoning) may simply be a mechanism subject to definitive description as a processor of syntactically-structured symbols. That makes it a mechanism embodying a theory--or nothing more than a theory. Heal 94 and the Davies/Stone NYU paper are pretty much the only useful sources directly on this point. So I suppose I'll have to take up there discussion next.

Haven't set a date for the 4th-year colloquium but I suppose I should pick something in late April or May.

A philosopher is a machine

A philosopher is a machine for turning Starbucks coffee into papers. Completed the S+B article draft today. There may be some changes down the line; who knows? Virgin Mobile may go out of business tomorrow, also.

The gears of the paper-producers

The gears of the paper-producers are beginning to creak. Prepared my first written sketch of the Strategy and Business article. Will share with editor this week. Begin refining draft for actual piece.