Saturday, January 19, 2013

A Memory from Grad School

Circa 1994: Josh Dever would be sitting on a couch in the philosophy graduate student lounge at U.C. Berkeley. I would propose to him a definition of "dessert" (e.g.: "a sweet food eaten after the main meal is complete"). He would shoot it down (e.g.: "but then it would be a priori that you couldn't eat dessert first!"). Later he would propose a definition to me, which I would shoot down. Over time, the definitions became ever more baroque. Other graduate students participated too.

Eventually Josh decided that he would define a dessert as anything served on a dessert plate. Asked what is a dessert plate is, he would say it was intuitively obvious. Presented with an objection ("So you couldn't eat Oreos right out of the bag for dessert?") he would simply state that he was willing to "bite the bullet" and accept a certain amount of revision of our pre-theoretical opinions.

At the time it seemed like cheating. In retrospect, I think Josh saw right to the core.

77 comments:

Neil said...

Important work, insofar as the correct definition of dessert is required for a complete theory of punishment. But I can't help wondering: what flavour were the bullets Josh bit?

Anonymous said...

in fact the definition of desert is great interest, of dessert it has no import

David Livingstone Smith said...

Funny, this is precisely the example that I use when trying to explain conceptual analysis to my Intro students.

Tamler Sommers said...

Why is it any less absurd to try to find the "correct definition" for desert, knowledge, responsibility, free will, table, chair, etc? I'll never understand that...

Neil said...

In favor of the claim that it is a priori dessert is always eaten after the main meal, I cite the argument from German. Auf Deutsch: Nachtisch.

Peter Hurford said...

What about "a sweet food *intended to be* eaten after the main meal is complete"

Current Berkeley Grad said...

A few years ago, some of us had an extended discussion about what a beverage is. Maybe there's something about Moses Hall that encourages these kinds of enquiries : )

Eric Schwitzgebel said...

Fun comments!

Peter: Intended by whom?

West. said...

Desserts are very often sour or savory (and non-sweet). Those of a Northern European persuasion love salty anise-inspired desserts.

Carlos said...

I just consulted my intuitions, and they said that Oreos eaten right out of the bag clearly count as a snack, not as dessert.

Eric Schwitzgebel said...

Carlos: As I recall, the dessert/snack distinction was frequently a bone of contention in these discussions.

Izzy said...

Of course there's no precise definition or "essence" of what constitutes a dessert, but rather to talk of dessert is to participate in a language-game of family resemblances and similar kinds. In this sense, apple pie is a dessert, but not in the very same way that a serving of cheese and crackers is a dessert.

nnyhav said...

alternatively ...

Callan S. said...

Intended by the average cultural practice of the culture Peter is part of, at the time stamp his comment was given.

Culture being a sample of a thousand people abducted at random, the average practice of intent, as measured by actions.

I'm not sure this is baroque. When you program, for example, you far more specific than this!

Francesco said...

The notion of a "main meal" is inherently vague and had better be avoided. And I don't know how to deal with the issue of (horrible) salty desserts. But of course it's analytic that a dessert is eaten at the end of a meal and cannot be eaten first. It's a matter of different ways of referring to the same thing: you can eat it first if you like, but then you don't call it a dessert. So a sentence like "It is possible for a dessert to be eaten first" has truth conditions analogous to those of "It is possible for a bachelor to be married".

Neil said...

Francesco: "dessert for breakfast" returns 821000 google hits. "dessert first" returns 974000. That's a heck of a lot of people who apparently can't use the English language.

JJ said...
This comment has been removed by the author.
dietl said...

X is a dessert iff (a) X is a object, and (b) there exists a person Y that calls X 'dessert'.

Richard Baron said...

We search for words that will define
Dessert, as eaten when we dine
But language is a desert where
No synonym is really there
Each answer but a mirage proves
This mad pursuit, we should not choose
But all as one, desert this text
Eat our main course, and ask what's next.

Eric Schwitzgebel said...

Richard, I'm sure you must have cribbed that poem from Wittgenstein!

Callan: Are you saying practices can intend? That sounds like a category mistake.

Dietl: So I can make the galaxy a dessert by calling it one?

Clayton Littlejohn said...

The nerve of Mr. Schwitzgebel's argument is that if A eats Oreos right out of a bag, then A eats a dessert without a dessert plate because a bag could never serve as a plate whatever one eats from it and regardless of whether one thereby eats dessert. While there may be much to be said for this view, still it does not seem right on the face of it.

Eric Schwitzgebel said...

(I might or might not have just called the galaxy a dessert. I'm not telling.)

Eric Schwitzgebel said...

Oh Clayton, we philosophers cannot rest on surface appearances. We must plunge to the creamy center.

dietl said...

hehe...okay
(c) X is edible.

dietl said...

oh, so for every Y that could eat the universe that's a 'Yes.'.

Neil said...

Dietl, I have printed out your comment. Mmmm, paper. Thanks for dessert!

dietl said...

bon appetite! hm...I prefer chocolate.

Richard Baron said...

Let's read McGinn, for he doth claim
That in the philosophic game
Truth by Analysis we find
Since each term can be well-defined
And this may compass every word
That at the table has been heard
From hors d'oeuvre to the final dish
For which in our life-form we wish.
We must all hope that this is true
Defining gives us work to do
And thus in tenure earn our bread
A poor dessert, and so to bed.

Callan S. said...

Eric, in as much as any process, like for example a self guided missile, intends a particular target by practice of smashing into it. Regular science already embraces that you might run an experiment 1000 times and get result A, but perhaps on the not run 1001 time, you might get result B.

Apart from that, in self guided aparatus, the practice indicates the intention. A practice can have an intention. It just has trouble conceiving of it that way around! >:)

Anonymous said...

Neil: "married bachelor" returns 40,900,000 hits on Google...

Neil said...
This comment has been removed by the author.
Neil said...

Anonymous: no it doesn't. You have added three zeros.

Argumentum ad googleum ftw.

dietl said...

What a vile thing to do. Manipulating the statistics just to prove your point ;-)

Scott Bakker said...

Dessert: Noise preventing toilet paper divination from becoming a true science.

The Ultimate Philosopher said...

Not my AOS at this time, but . . .

Merriam-Webster's proferred definition is "a usually sweet course usually offered at the end of a meal." Is a definition wishy-washy if it employs "usually" (not just once, but twice)? Is there a reason a definition could not employ "usually"? It might be asked whether we're talking about the "merely general" (an "always or for the most part") or a linguistic universal, but since when couldn't even a universal employ "usually"s? It covers any and all instances of our use of the term "dessert," doesn't it? I dunno, I'm just an amateur on this subject at this point. -UP

Polo said...

Might I remind Mr. Schwitzgebel that Galactus, Marvel villain, consumes entire planets and other planetary constellations. Let me revise deitl's statement:

So if "X is a dessert iff (a) X is a object, and (b) there exists a person, or Marvel villain, Y that calls X 'dessert'."

One could imagine Galactus saying "After consuming every planetary constellation that comprises Galaxy X, I will, then, eat planet Earth for dessert" or something to that effect.

Joshua Harwood said...

Yes, normative armchair lexicography is a fun, yet empty intellectual exercise.

But what areas of philosophical interest and investigation aren't like that?

dietl said...

@Joshua:

Can you tell me a non-empty intellectual exercise?

Joshua Harwood said...

@dietl

Sure, devising hypotheses for which one can provide empirical tests to confirm or disconfirm them isn't an empty intellectual exercise.

dietl said...

What makes this non-empty? The "empirical"?
What makes Math non-empty? Is it?

PS: Not trying to be annoying, I only want to know what/how other people think :-)

Joshua Harwood said...

@dietl

My litmus test for the emptiness of a given discourse concerns the output that such a discourse provides.

If a discourse on a topic directly contributes to an empirical discovery that improves technology or living standards, then I would say that it's non-empty.

If a dialogue just results in more dialogue, and people can't even agree on what kind of test would falsify the discussed claim, then I'd say that it's empty.

Most math and logic passes this litmus test. But philosophy of math and philosophy of logic...?

I think that people who want to ponder, say, modal realism, in exchange for a salary owe us some empirical results demonstrating that they're not wasting intellectual and financial resources.

My motto for this can be: "Games are fine as long as I can choose whether to pay you to play them."

Anonymous said...

Joshua Harwood writes that "If a discourse on a topic directly contributes to an empirical discovery that improves technology or living standards, then I would say that it's non-empty.[...] Most math and logic passes this litmus test."

But a great deal, perhaps even most, math and logic clearly does NOT pass this litmus test, since a whole lot of formal logic and higher mathematics is theoretical and not applied: much of it has no empirical use value and is of interest only to other specialists.

Even in the physical sciences, the demand for practical and empirical use is not much help: the history of human invention is such that we do not always or even usually know in advance what practical uses all scientific experimentation will yield.

The appeal to knowledge which only "directly contributes to an empirical discovery" seems narrowly dogmatic, and assumes a kind of hindsight that even the physical sciences cannot attain.

Neil said...

Joshua: nice philosohical theory you have there.

dietl said...

@Joshua
I think that your "litmus test" is a good start for you definition of empty. But apart from other formal problems I would say that it is a bit too hard for not all of math and logic passes the test. Also history, ethics and other disciplines that I would describe as "cultural achievements" might fail your test. But that's all debatable.
One thing that bothers me is that you shouldn't make anything depend on people agreeing on anything, because this might prove to be a fatal criterion ;-)

If being empty makes a discussion valuable or not and if it makes it worth paying for, that's another thing that's debatable and depends on economic factors. If a country hasn't enough money to warrant putting a lot of it into scientific research, that doesn't make it unvaluable. So its also about who can choose to pay for it.

Joshua Harwood said...

@ Anon 1/21

I would contend that most logic passes this litmus test, especially since I spend considerable time on the prospect of translating all natural language utterances into some form of higher-order logic, which then can be used to improve natural language user interfaces and natural-language inference programs.

Mathematical and logical advancement has generally been a required precursor to more accurate interpretations of empirical tests' results and to experimental modeling, but I could bite the bullet in certain areas of number theory where the results appear to offer no help to any foreseeable benefit to any hard science project.

If one can accurately translate an empirical prediction reliably under a formal language, that generally strengthens the reliability of the claim. (e.g. Proving the rate at which the temperature of a body increases as it approaches the sun greatly strengthens the truth of the claim, "Things that approach the sun get hotter.")

As for the history of human invention, I would remark that abandoning empirical projects upon the falsification of certain hypotheses has been immensely helpful. We learn to be more correct in large part by learning to be less wrong.

I appreciate the input.

@ dietl

The "cultural achievements" that you mention do fail that litmus test. Living standards have improved as a result of certain "revolutions," but they have equally worsened under them, and I don't think that a historian or ethicist has any greater insight to impart regarding why Gandhi's project of liberating India from British subjection was good, while Hitler's project of ethnically cleansing the world was bad.

If we organized societies based on empirically rigorous accounts of something that falls under the phrase "human nature," such that we sought only to maximize the satisfaction of human preferences within the natural bounds of scarcity, I don't see any further need for musing on its rightness or wrongness.

dietl said...

Imagine a world where noone would make historical investigations. Maybe one day Hitler and his politics would be forgotten and noone would remember the damage that had been done. Who but a ethicist could tell the world what is right and wrong ;-) Of course I'm exaggerating with this.

I think we can agree to disagree on your first paragraph ("The "cultural achievements"...") but the second one might sound great at first but everyone with a deeper understanding of ethics can tell you that such a principle raises more questions than it solves. By the way, have you read "Brave new world"?

Joshua Harwood said...

@dietl

Sorry, my dispositions make tolerating a novel nearly impossible. I like brief stories, though. Are there any fables, shorter allegories, dialogues, or thought experiments, or clear arguments which get to whatever problems you think that novel makes more curtly?

I'm sure that this will present an opportunity to work on my "depth."

Anonymous said...

Joshua Harwood writes "If we organized societies based on empirically rigorous accounts of something that falls under the phrase 'human nature,' such that we sought only to maximize the satisfaction of human preferences within the natural bounds of scarcity, I don't see any further need for musing on its rightness or wrongness."

1) Who is the "we" in the first sentence meant to refer to? Whomever it is, on what does their power rest? Is it a committee, a directorate, an aristocracy of philosopher kings, a Mohist council, what?

You seem wholly unaware of how quickly you have entered some very old problems in political philosophy: a discipline you likely scorn in your narrow zeal for so-called empirical results.

2) There is no agreement on whether there is such a thing as human nature, let alone any uncontroversial and "empirically rigorous accounts" of it. It's not even clear what an account would consist of. The debate about this is very old and yet you give no indication that you are familiar with it in any way.

3) Your political vision strikes me as a paternalistic and naive form of totalitarianism masquerading as utilitarianism.

Some indication of familiarity on your part with the kinds of humanistic philosophy you cavalierly dismiss (such as ethics and political philosophy) might be helpful.

dietl said...

Hm, I suggest reading the wikipedia page for an overview, for I can't think of another story right even though there might be a lot of similar books.

The point I wanted to make is that "maximize the satisfaction of human preferences" leaves a lot of room for interpretation.

I see now that Anonymous already wrote something similar to what I wanted to say, so I just have to agree with him.

Joshua Harwood said...

Hi, Anon (I guess the same one):

I post my full name and every post here links to my Blogger profile, which links to a blog where I consider arguments and takes from various sorts. I don't know who you expect me to quote mine or namedrop into this textbox to demonstrate a certain amount of relevant knowledge.

Political power is the ability to direct people to certain behaviors and actions. How one accomplishes this differs in specifics, but has some general trends.

I don't have any burden to say exactly who will have that power (maybe a king and lords, maybe an autonomous collective), so your (1) won't be relevant to me, and I leave it to empirical findings to flesh that out.

But I think that you'll discover quickly that I'm far from a paternalistic totalitarian. In fact, some (who read my blog from time to time) have accused me of being exactly the opposite.

@dietl

Yes, I'm aware of the problems in subjectivity, the problem in finding any universally true facts about humans that would reliably stick into any informative model on how to guide behavior toward what makes everyone "happy."

At the same time, though, I think that (a) there may be shortcuts around these problems, and (b) that such shortcuts could actually lead to more research that makes this foggy notion of "human nature" clearer, and thus make incentivization and prohibition less error-prone.

dietl said...

I hope this is not too much (or even rude?)to ask for but I would really like to know what "dispositions" you were refering to earlier. I read some of you blog posts and this might help to give me a more complete picture. From what I've read so far I would conclude the following:

-You know how to use logic.
-You have a quite shallow understanding of the problems of philosophy.
-On your blog you criticise some philosophical statements and while you sometimes make good points at other times your critique fails because of misinterpretation.
-You are a philosopher.

On the last point you might disagree due to a different understanding of 'philosopher'. It seems to me that what you mean with it is the job(?)/ profession of being a philosopher. Someone who gets paid for it and claims to be an "expert". But what I mean with it is just someone who is interested in asking questions.
I think you would greatly benefit from going a bit deeper into philosophical texts. If you want to have any suggestions you only need to tell me what topic interests you. The only thing that might be a problem is what the question from above is about: your "dispositions".

Anonymous said...

Joshua Harwood: it's not name dropping or references I'm after so much as better reasoning and more humility, i.e. perhaps an admission that one should be skeptical of platitudes and one-size-fits-all proclamations about something as difficult as how best to organize the ideal society.

You write that "political power is the ability to direct people to certain behaviors and actions," but I'm unsure what is political in such a definition. The maitre d' at a restaurant has the ability to direct people to certain behaviors and actions, but the power he or she has does not seem especially political.

You then write that "how one accomplishes [the execution of political power] differs in specifics, but has some general trends," yet fail to specify what those trends are.

You may think you "don't have any burden to say exactly who will have that power," but this is a little like leaving a recipe without any ingredients.

Furthermore, I find it hard to believe that you really do think something as vague as "empirical findings" will "flesh...out" the details of the ideal political organization. Whatever would lead one to expect that there's the potential for conclusive empirical findings for such a thing? It strikes me as patently absurd.

You also dodge the ethical question: if your much vaunted "empirical findings" deem that state coercion, deception, torture, and political manipulation are the most efficient way to maximize "the ability to direct people to certain behaviors and actions," is such power justified?

Joshua Harwood said...

I'm sorry, Anonymous (if you're the same one), but you've succumbed to a few straw men, forging an ethical "problem" from a sentence that I did not write:

I never wrote, "maximize 'the ability to direct people to certain behaviors and actions'" implicitly or explicitly, anywhere. CTRL+F is your friend.

You did this in the same post that criticizes me for not attending to the details of what I claim.

I guess it was a wise move for you to remain anonymous.

Joshua Harwood said...
This comment has been removed by the author.
Joshua Harwood said...

@dietl

I don't mind telling you my disposition.

I can't enjoy novels because I can't stand to read overlong make-believe narratives and quotes from mouthpieces who, in fictional settings where novelists control the imagined circumstances of their mouthpieces, steer their own fictitious plots to confirm whatever arguments they prop up via those mouthpieces. I don't mean to muddy that with so many clauses, but that's what makes novels unbearable.

That, and they're too wordy.

It's not mere fiction that bothers me. Zhuangzi discusses Liezi's legendary ability to float on the clouds, and then states that people falsely assume that he is "真正自在" ("truly self-subsisting"), because he still relies on the clouds to float. But we can see the premises to his arguments clearly. Also, it's easy to substitute his fanciful analogy with a concrete one which lends evidence to its truth.

Daniel Dennett made some comments about William Lane Craig that I think capture an analogous distrust that I have of most philosophers who I've read. (I reposted this to get the URL to work.)

dietl said...

It seems to me that you fear that a novelist, just as a philosopher is trying to trick you into believing something. But that's not the point of philosophy or of a story.
One thing you have to keep in mind is that philosophy and other disciplines are academic battlegrounds. Every theory gets "attacked" by other philosophers and the things you criticize most likely have already been criticized by other philosophers. With philosophy you don't learn to convince people but to avoid mistakes. I'm not saying that you have to believe every one of them, but you don't seem to consider that they might be right and dismiss them to fast.

In your post "Philosophical Amputations 2" you write the following:
"Not too long ago, a philosopher handed me some sample questions and asked me whether empirical science could answer them better than philosophy could.(...)

1. Are there any genuine ethical obligations beyond the obligations to maximize the happiness and preference satisfactions of present or future beings?"

In this post you deconstruct this question with you logical methods but you fail to answer the original question (because you don't want to answer it?). Anyway, what did is show (without realising?) what the better answer is.

It's quite ironic, that every post you label anti-philosophy contains no "anti", but a lot of philosophy.

Joshua Harwood said...

@dietl

I figured that the original question, whether empirical science could answer that philosopher's questions better, had an obvious answer:

No, questions with unjustified basic assumptions (in Amp-2, terms without demonstrated referents) cannot be investigated by empirical means or mere argumentation. Philosophers can't answer the question that I used in the example, and neither can scientists, so in terms of the probability of satisfactorily answering such questions, ~(0% > 0%).

If someone wants to argue about concepts like "goodness," or "obligations," or whatever, he'll have to point to measurable or reproducible instances of them, first. He has to make the terms and predicates of his language code for things that we observe, or at least could observe in a controlled setting.

I use sentences like that philosopher's to provide data points to demonstrate that most of the work of philosophers is strictly nonsense, and what isn't nonsense is already a part of the standard practice of empirical sciences, mathematics, and maybe economics, which leads me to question the legitimacy of calling any academic exploration particularly "philosophical."

Is my answer above not the answer that I showed?

dietl said...

"If someone wants to argue about concepts like "goodness," or "obligations," or whatever, he'll have to point to measurable or reproducible instances of them, first. He has to make the terms and predicates of his language code for things that we observe, or at least could observe in a controlled setting."

That's your dogma. I don't think you can point to measurable data that support this claim.

"most of the work of philosophers is strictly nonsense, and what isn't nonsense is already a part of the standard practice of empirical sciences, mathematics, and maybe economics,..."

How do you know what "most of the work of philosophers" is, if you complain about novels being to wordy? I suspect you didn't read most of the work of philosophers. You make it pretty easy to dismiss philosophy if you define everything that philosophy does that isn't nonsense as belonging to other disciplines. I think if you had a deeper understanding of the roots of empirical sciences you would have a bigger respect of philosophy.

"Is my answer above not the answer that I showed?"

How did you get to your answer? trough philosophical reasoning. So saying "Philosophers can't answer the question" is wrong. A different answer than what the question suggest is not no answer.

Joshua Harwood said...

@dietl

You write: "That's your dogma. I don't think you can point to measurable data that supports this claim [that terms and relations must code for things that we could observe in order for us to evaluate their truth]."

I think that's a fair challenge. Empirical data to show that you must be able to meaningfully refer to the things in questions in order to evaluate their truth (beyond vacuous truth) is straightforward enough. Consider this model. Introduce this question to subjects:

Q: 現在的法國國王是不是光頭的?

Now, we can wait until someone actually performs this test, but here's how I suppose it will play out (with subjects John and Wang).

Let's assume that John can't read Mandarin. He doesn't have a clue what the sentence reads, but can guess that it's a question because he can identify the "?" in the sentence.

He can't match any of the terms or predicates to any things that he's observed or not observed, and so simply can't provide a coherent answer to the question. It's strictly meaningless to him.

Wang, however, does read Mandarin (and translates English in the parentheses). He phones his French Embassy to discover that France is a Republic, and that there is no "現在的法國國王" ("present King of France").

Next, teach John and Wang the relevant fundamentals of logic and set theory.

Wang can now illustrate the problem. The present King of France codes for all of the things that are in the empty set -- ∅. Therefore, whether "__是不是光頭的?" ("Is __ bald?") is vacuously true, because the false antecedent allows for the truth in both answers.

We get:
(x∈∅)⇒(P(x))
(x∈∅)⇒¬(P(x))
¬(x∈∅)

Some other arguer can try to change the semantics of '⇒,' which has been done, but not in a way that overturns this matter, or you can opt to negate (or if you allow DNE in your logic, un-negate) one of those premises. Good luck to him!

The philosophical question that I dissected is not different from this. Philosophers have yet to show that the elements of "obligations" label one observable element. Once they do, they can have their battles, except they will be redundant, because at that point scientists could do the research more quickly and efficiently once technology and resources allowed them to investigate there.

You also write: "How do you know what 'most of the work of philosophers' is, if you complain about novels being to wordy? I suspect you didn't read most of the work of philosophers."

Your suspicions are correct if you mean that I haven't read over 50% of all of the books, articles, etc. that philosophers associate with their profession. But neither has anyone else.

However, I do know enough about the ingredients of "philosophical" argumentation (by reading plenty of philosophers' books, some of which are too wordy) to show where to undercut them and deny them relevance or import.

Joshua Harwood said...

You also write: "You make it pretty easy to dismiss philosophy if you define everything that philosophy does that isn't nonsense as belonging to other disciplines."

That's an empirical matter, not a definitional one. I can walk into any department and find people arguing rationally about their arena of expertise, even publishing those arguments to recommend directions of empirical research, establish hypotheses for testing, etc.

You write: "How did you get to your answer? through philosophical reasoning."

I got to my answer through an argument and through a fixed method which uses (self-?)proclaimed philosophers' sentences as survey data. None of the methods nor any particular conjunction of those methods is unique to philosophical study.

I feel that it's an unfortunate fluke that logic is still housed in philosophy departments, when it really belongs in the mathematics and computer science departments at this point, seeing how most of the greatest logicians of the past hundred years were/are mathematicians and computer scientists, not philosophers.

dietl said...

"That's an empirical matter, not a definitional one."

I have to disagree with you here, but I see no way and no reason to convince you now since you seem to be stuck in you thinking about what philosophy is.

"I feel that it's an unfortunate fluke that logic is still housed in philosophy departments, when it really belongs in the mathematics and computer science departments at this point, seeing how most of the greatest logicians of the past hundred years were/are mathematicians and computer scientists, not philosophers."

I think you have lot of research to do concerning the history of sciences. I'm not here to lecture you on this, but if one day you want to stop making up facts to create you own reality I would definitely suggest reading some wikipedia articles about the topic for a start, even though they might be too wordy.

dietl said...

You write:
"You write: "That's your dogma. I don't think you can point to measurable data that supports this claim [that terms and relations must code for things that we could observe in order for us to evaluate their truth].""

Sorry, but I don't think I meant want you wrote here. What I wanted to say is the following:
I don't think you can point to measurable data that supports this claim [that if someone wants to argue about concepts like "goodness," or "obligations," or whatever, he'll have to point to measurable or reproducible instances of them, first].

Which measurable data makes sentences like 'x is good' true? Furthermore which measurable data makes the sentence 'Measurable data is needed to argue about sentences like 'x is good'.' true?

Joshua Harwood said...

@dietl

Do you really not have a way, or do you not want to attempt to tell me a way? I have a lot of arguments, and backhanded suggestions from unestablished authorities don't counter them.

You ask: "Which measurable data makes sentences like 'x is good' true?"

You can start by showing what you mean by the word 'good', which most people do by giving extensional cases.

"Eight hours of sleep is good." (healthful?)
"Delicious food is good." (enjoyable?)
"Coen Brothers movies are good." (entertaining?)
"Charity is good." (helpful?)

Then, whatever you mean by 'good' is whatever those things have in common. Thus far, if I had to take all of those into consideration, I would surmise that "x is good," means something like many people seem to want x (but maybe you've read my blog post on this).

But we could play this kid's game with any term: How do we know that "x is blue," is true?

First, provide some extensional cases to train him to associate the phoneme or grapheme 'blue' with those cases, have him roughly guess at some definiens (somewhere in the intersect of all of those things that you're training him to refer to as "blue").

"Bluejays are blue."
"The sky is blue."
"Blueberries are blue."
...

But how do we measure whether "x is blue" is true? Well, the measure for truth is usually probabilistic or Boolean.

To get those measurements, you round up a sample of native English speakers, instantiate those things x that you're checking for blue, and map out the range of shades of colors that people are statistically likely to say are blue.

Then the statistically relevant intersects of measurable features are what 'blue' is for English speakers at the time of that study, and it's as blue as blue-ness is.

Disappointing? Think how I feel to watch philosophers agonize over this stuff.

Your own comment on Sun Jan 20, 05:19:00 AM PST suggests that you actually agree with my approach for the most part.

Or we could not do any of that, and just hand-wave at the "Form of the Blue," or whatever jargon gets your gander, and get no closer to an answer. I prefer desert landscapes.

Joshua Harwood said...

I should clarify and remark that "blue" is a term, in the sense "any word or group of words considered as a member of a construction or utterance," but isn't a term in the logical sense, which we usually designate to mean a constant or variable which serves as the argument within an ordered n-tuple in the construction of atomic sentences.

In the latter case, "is blue" is a predicate.

dietl said...

"To get those measurements, you round up a sample of native English speakers, instantiate those things x that you're checking for blue, and map out the range of shades of colors that people are statistically likely to say are blue."

What follows from your reasoning is that a thing x is blue relative to what different statistics say about it. So it is possible for an object x to be blue at a given measurement and not blue at another.

Also this:
"instantiate those things x that you're checking for blue"
What if all the things you provided for checking are not only blue, but also share a triangular form.
Then if you say "x is blue", people might think that you were refering to a blue and triangular object.

Okay enough with the "kid's game".
Which measurable data makes the sentence 'Measurable data is needed to argue about sentences like 'x is good'.' true?

Joshua Harwood said...

I think that I need to chase you to the end of the rabbit hole on the first line of questioning before I tell you the answer for the second. If all and only blue things were triangles, and it were impossible to devise an instance of non-blue triangles (e.g. with paint) or blue non-triangles (e.g. with hammers), especially if the linguistic community insisted on it, then "blue" means just that.

Just out of curiosity, how many languages do you speak? How many do you think you would speak if you couldn't reliably trust even the most basic terms' meanings?

Your phrase of the day should be "à la mode."

If you can stand the sight of blood, I can go on to show you that not all and only mammals that have hearts have kidneys.

dietl said...

" If all and only blue things were triangles, and it were impossible to devise an instance of non-blue triangles (e.g. with paint) or blue non-triangles (e.g. with hammers), especially if the linguistic community insisted on it, then "blue" means just that."

So you think it would be impossible to distinguish between the form and the colour of a "blue" object? Even if this distinction would be no problem with other colours?

I speak two languages fluently and I don't think I need to reliably "trust even the most basic terms' meanings" of a language to be able to speak it. Communication only has to work, it doesn't need to make sense. So the answer to the second question is also two.

Joshua Harwood said...

@dietl

"So you think it would be impossible to distinguish between the form and the colour of a 'blue' object?"

Why do you think such an imagined state of affairs implies that? If in Wc (counterfactual world) all and only blue things were triangles, "blue" in Lc (counterfactual language) coded them, and if people could not alter that state of affairs, or if the community refused to let "blue" in Lc refer to just blue things like Wa and La (actual world and language), then the predicate "is blue" in Lc means "is a blue triangle" in La (actual language).

"I don't think I need to reliably 'trust even the most basic terms' meanings' of a language to be able to speak it. Communication only has to work, it doesn't need to make sense."

Communication only works when people agree on a language, which they do so that the messages will make sense for them. Language that doesn't make any sense to someone doesn't work.

I speak three languages proficiently, so if you don't read Mandarin or Spanish fluently, I can give you plenty of messages that won't make sense or work, since you won't be a member of at least one community for whom their language's sentences do make sense.

"So the answer to the second question is also two."

And yet you don't suffer this problem?

dietl said...

Could it somehow be possible in Wc to define a predicate P which in La means 'x is blue'?

"Communication only works when people agree on a language, which they do so that the messages will make sense for them. Language that doesn't make any sense to someone doesn't work."

Languages as a whole don't need to make sense, by which I mean that they don't need to be logically coherent, in order to work. I can speak English with a child even if neither me nor the child agreed on a language. I say 'Give me the ball.' and the child hears the words 'give', 'me', 'the', and 'ball' and knows how to respond to them. Where does anyone agree on something here?
Noone needs to know the meaning of a words to communicate, only how it is used. Knowing the meaning of a word is good but not necessary.

"And yet you don't suffer this problem?"

That's a problem where your definitions lead to.

dietl said...

"That's a problem where your definitions lead to."

I hope you don't take this literally, it was meant more like: "Me? No YOU!" ;-)

Joshua Harwood said...

"Could it somehow be possible in Wc to define a predicate P which in La means 'x is blue'?"

Since Wc is just some world that you're imagining and that I'm entertaining, in principle, you could illustrate this world such that it is impossible.

But assuming that, ceteris paribus, the facts of Wc insersect with the facts of Wa, except that all and only blue things be triangles in WC, then yes, because there will still be red, black, yellow, etc. things.

If there were Lc-speaking blind people in Wc, then "is blue" would just mean "is a triangle" to them.

If there are certain scanners in Wc which don't allow for color resolution, then Wc is only macroscopically different from Wa, and there would be a way to see things such that they were triangles, but not blue.

Those two data sets would suffice to show that we could describe Wc's so-said "blue" things with two distinct predicates. They could even use the x-eme "blue" to refer to it, as well.

You say: "Languages as a whole don't need to make sense, by which I mean that they don't need to be logically coherent, in order to work."

Can you name or construct a logically incoherent language that works?

"I can speak English with a child even if neither me nor the child agreed on a language. I say 'Give me the ball.' and the child hears the words 'give', 'me', 'the', and 'ball' and knows how to respond to them. Where does anyone agree on something here?"

First, unless you have presented a child with some initiation rituals to get children to guess the meanings of utterances, the child will have no idea what a "ball" codes. Go to a kid who doesn't speak Mandarin and only command, "給我求!" over and over, miles away from the nearest ball, and you'll see just how ineffective it will be.

Agreement doesn't have to be explicit, but sometimes it is.

In fact, one driving hypothesis of language acquisition is literally "comprehensible input." Here's the main guy who proposes that hypothesis illustrating exactly how you need to indicate references so that people will make sense of your utterances.

You may notice how the audience didn't start questioning him in his lesson or react to him as if he were wrong. That's because they were accepting his claims at face value, accepting his authority in the language to teach it to them.

Children are similarly slavishly obedient.

Joshua Harwood said...

There's a typo in the Chinese sentence. It should read, "給我球!"

dietl said...

"Since Wc is just some world that you're imagining and that I'm entertaining, in principle, you could illustrate this world such that it is impossible."

Do you think that's no problem? That there might be a world where it is impossible to define such a basic word. Again, you said this is kids game and still you seem to struggle. Shouldn't you maybe think over you methods for definition? On this topic I would suggest for you to read some Russell, Carnap or Quine. You don't really seem to think about the consequences of what I'm saying and aren't open to the possibility that you aren't right.

"But assuming that, ceteris paribus, the facts of Wc insersect with the facts of Wa, except that all and only blue things be triangles in WC, then yes, because there will still be red, black, yellow, etc. things."

That's not what we were talking about. All things that are blue are also triangles, not "all and only".


"Can you name or construct a logically incoherent language that works?"

Any language where the liar paradox is possible. I'm not sure logically incoherent was the right word. But that's what I meant with it. Here my suggestion would be Tarski and Gödel.

"In fact, one driving hypothesis of language acquisition is literally "comprehensible input.""

Thank you for helping me prove my point. See what I wrote before:

"Noone needs to know the meaning of a words to communicate, only how it is used. Knowing the meaning of a word is good but not necessary."

The people who get comprehensible input don't know the meaning or defintion of a word they only guess at it and it still works.

Are we finally ready for this or do you want to continue sidetracking?
Which measurable data makes the sentence 'Measurable data is needed to argue about sentences like 'x is good'.' true?

dietl said...

"There's a typo in the Chinese sentence. It should read, "給我球!""

No problem I can't read it anyway on my computer and if I could I wouldn't mind ;-)

Joshua Harwood said...

"Do you think that's no problem? That there might be a world where it is impossible to define such a basic word."

No, it's not a problem. I've taken you through the steps to show you how it's not a problem.

"On this topic I would suggest for you to read some Russell, Carnap or Quine."

Christ on a cracker! You apparently haven't noticed how I've been dropping quotations and paraphrases from the likes of those aforementioned people, latter-Wittgenstein, Austen, Putnam, Rorty, and so forth.

Maybe we should take some time to discuss Quine's piss-poor comprehension of lexicography (in "Two Dogmas...") or failures in comprehending modal logic. Don't assume that I'm illiterate on this topic because I argue that it's empty.

"That's not what we were talking about. All things that are blue are also triangles (in Wc), not 'all and only.'"

It doesn't matter, I've shown that Wc allows for P's for Lc that in La mean "non-blue triangles," and there are just as many where I can show that Wc would allow P's for Lc that in La that mean "blue non-triangles." (Some people in Wa cannot discern shapes correctly, but can see colors. Also, if it's at all possible to magnify into blue triangles in Wc without actually hurting the ability of it to reflect color, then they could have a predicate for blue_La things.)

"I'm not sure 'logically incoherent' was the right word. But that's what I meant with it."

Neither Tarski nor Gödel argue that Classical logic or Peano arithmetic was logically incoherent. Incompleteness is not a proof of total logical incoherence. An indefinability theorem is not a proof of total logical incoherence. Now I'm beginning to doubt whether you really understand the writings of the authors that you namedrop.

And if you had read your Tarski, you would have known that he's proposed a very adequate solution to the liar's paradox, as many others have done.

"The people who get comprehensible input don't know the meaning or definition of a word they only guess at it and it still works."

That is knowing the meaning of a word.

Amazingly, in one sentence, you've only demonstrated a complete lack of knowledge of the empirical science (specifically in lexicology and cognitive semantics) which says that you're dead wrong on this front. How many links do you need to set you straight?

Here's a first hint: Knowing an exact definiens to any term, given semantic shift and the ability to infinitely analyze and coin new terms for new observations, is not feasible, and knowing the meanings of terms in natural languages has never, ever required that level of precision.

Your issue apparently arises from a severe misunderstanding of language. Sound like anyone you could namedrop?

"Are we finally ready for this or do you want to continue sidetracking?"

Sidetracking? I'm going to reduce your nonsense here to oblivion first, and then I'll bury you there. There's no rush for me.

I'm losing my patience, though, so I'll give you one chance to admit that you're wrong here, and then move on to the next horn, or you'll just have to guess at the answer for yourself.

dietl said...

Actually I don't know where to begin. you last post reads pretty weird to me. You write sentences that are from my perspective complete nonsense and still you say that it is me who talks nonsense. You don't really listen to what I'm talking about you only go on talking about things you assume to be right.

"That is knowing the meaning of a word."

No, this is knowing the use of a word. And this makes pretty clear that we define things very different.

"Sidetracking? I'm going to reduce your nonsense here to oblivion first, and then I'll bury you there. There's no rush for me.

I'm losing my patience, though, so I'll give you one chance to admit that you're wrong here, and then move on to the next horn, or you'll just have to guess at the answer for yourself."

The only thing you reduce here is my faith in your intelligence, even though your blog gives me reason enough for this. You know what the funny thing is? Even if you were right and I would be wrong in this special point you STILL keep sidetracking and even begin to put ridiculous words in my mouth I didn't say. There is no reason for not answering my question except that you don't WANT to answer it or can't.

dietl said...

I know it's very hard to comrehemd for someone like you but still here is a useful link:

http://en.wikipedia.org/wiki/Principle_of_charity