Thursday, May 24, 2007

Modal Copulation and Satisfaction. by Jessica Leech

Many philosophers, for various reasons, feel uncomfortable with the use of possible world semantics in an explanation of modal language (fragments of natural language including modal locutions such as ‘possibly’, ‘necessarily’, ‘must’, ‘could be’, etc.). A common alternative suggestion is that we analyse modal statements in terms of modal predicates and/or modal properties. For example, “Socrates is necessarily a man” is understood as something like predication of ‘being necessarily male’ of Socrates. An amendment, and rival, of this approach understands the modal term, not as modifying the predicate, i.e. from ‘male’ to ‘necessarily male’, but as modifying the copula, i.e. from ‘is’ to ‘is necessarily’.

This kind of approach adheres to the intuition that de re modality is somehow involved with the relation between an object and a property, rather than being involved in the inception of new predicates and new properties. This can be illustrated by the difference between asking, ‘Is Socrates necessarily wise, or is he just contingently so?’, and asking which of two properties Socrates has, (contingent) wisdom or necessary wisdom?

McGinn, in his book Logical Properties, presents such a copula modifier account of modality. He claims that a statement of the form “a is [modally] F” is true just when a [modally] satisfies “is F”. So, for example, “Socrates is necessarily wise” is true iff Socrates necessarily satisfies the predicate “is wise”. Modal satisfaction is in turn explained in terms of modal instantiation. So, Socrates necessarily satisfies “is wise” iff Socrates necessarily instantiates wisdom.

The challenge for modal copula accounts that employ “modal satisfaction” is to flesh out this new semantic notion. McGinn’s strategy is to appeal to modes of instantiation. He claims that there are two and only two modes of instantiation, necessary and contingent. Our statements might leave this as neutral, e.g. “Socrates is wise”, but at the worldly level, where we find Socrates and wisdom, if they bear the instantiation relation to each another, then they bear that relation in one or other mode.

The problem with this theory is that there is no way to accommodate mere possibility. Firstly, it is prima facie unclear how one could begin to understand possible instantiation as a mode of instantiation, where there need be no actual instantiation, but there is only mere possibility. More worrying, a contradiction can be generated:

(1) “Socrates is possibly foolish” is true.
(2) So Socrates contingently instantiates foolishness.
(3) So, Socrates instantiates foolishness.
(4) “Socrates is not foolish” is true.
(5) So Socrates does not instantiate foolishness.

In order to give the copula modifier account of modality fair trial, we cannot follow McGinn’s lead. We must either find an alternative way to understand modal satisfaction, or find a better way to make sense of copula modification than in terms of modification of satisfaction.

Labels: , , , , , ,


Anonymous Anonymous said...

As I understand it, logic is a tool for demarcating the difference between valid and invalid statements, or as Wittgenstein saw it, to

demonstrate what we can and cannot speak of. Illogical statements are meaningless, and should therefore be thrown out.

This enables the philosopher to filter out nonsense and pursue meaningful discussions.

Modal logic, being an extension of plain old Logic, goes further by bringing in more metaphysics: the notions of possibility and necessity.

What then, is modal logic good for?

Well, Boolos (1993) argued that it has proved useful for central results concerning provability in the foundations of mathematics.

However, mathematical considerations are somewhat different to those discussed in the original post: possible world semantics.

View #1. Modality concerns a relation between an object and a property. Viz., an object can have a property/predicate necessarily or


Consider the following Non-modal statement Re. Socrates:

"Socrates is a wise man".

This can be broken down to focus on the two predications:

"Socrates is a man AND he is wise".

Utilizing the two modal operators, "necessity" and "possibility", we derive the following modal statements:

Modal #1.

"Socrates is necessarily wise and he is necessarily a man".

Modal #2.

"Socrates is contingently wise and he is contingently a man".

Modal #3.

"Socrates is contingently wise and necessarily a man".

Modal #4.

"Socrates is necessarily a man and contingently wise".

As we have two predicates,"wise" and "a man", and we have two modal operators, "necessity" and

"possibility", we have 2x2 = 4 possible permutations of the original sentence in modal form, listed above, labelled #1-#4


Here are the first two problems with the above #1-4 modal statements above:

1) The properties of being a man and being wise are metaphysically different: being a man seems to be using the

"is" of identity, rather than the "is" of predication: Socrates IS a man. Full stop. Whereas the predicate of being wise

seems more like the "is" of predication: Socrates has the property of being wise.

This hits upon a central problem of modal statements:

We can surely say all predicates, where we use the "is" of predication, are contingentstatements: there is nothing to entail

that it is necessarily the case that Socrates IS wise, unless of course, we build the predicate "wise" or "wisdom" into our

definition of Socrates himself.

But Socrates is necessarily a man: if he were a fish, for a example, or a mongoose, then we'd be having a very different


McGinn's account of "Modal Satisfaction" seems far from satisfactory.


Well, on a prima facie reading it seems to be identical to ordinary logic's notion of satisfaction.

Remember Tarski's old sentences: "Snow is white, iff, Snow is white".

McGinn writes:

a statement of the form “a is [modally] F” is true just when a [modally] satisfies “is F”. So, for example, “Socrates is necessarily wise” is true

iff Socrates necessarily satisfies the predicate “is wise”. Modal satisfaction is in turn explained in terms of modal instantiation. So, Socrates

necessarily satisfies “is wise” iff Socrates necessarily instantiates wisdom.

Do you see a discernible difference between McGinn's and Tarski's?

I certainly don't...

Returning to my pedantic four modal statements derived from the original statement, "Socrates is a wise man":

The problem here centers around what sort of predicates/properties we're modally assigning to Socrates.

The name "Socrates", for example,refers to him. It is trivially true that it is contingent that Socrates is called "Socrates". Surely it

is possible that he was named "Stavros".

But as the discussion is about Socrates, we must choose which elements to hold fixed and which to examine metaphysically.

Let me explain by way of analogy:

In any scientific experiment, one needs variables and constants: the constants give us the frame of reference, and

the variables are what are under examination/scrutiny. We need something fixed by which we may have a bench mark to compare the


Well, the same principles apply to equally to modal logic where we're comparing different properties/predicates of an entity [be it a

person, object, city, etc]. We need something held fixed while we scrutinize the modality of other properties.

Consider the sentence once again: "Socrates is a man and he is wise".

We have "Socrates", "is a man" and "is wise".

Socrates is the subject, and "is a man" and "is wise" are the predicates.

So we have to keep "Socrates", the subject, kept constant while we explore the modal possibilities of him being a man and being


So we may permute the sentence using the two modal operators of necessity and possiblity and derive the four sentences listed earlier

in this post.

But if "is a man" is deploying the "is" of identity rather than the "is" of predication, then by changing that, we're

changing what it is for Socrates to be Socrates.

This is an illegal move in the modal game.

So we're only left the predicate "is wise".

We may investigate this modally, as it is a property because we're using the "is" of predication, not identity.


What does all this have to do with the problem stated in the original blog post, of finding a satisfactory account of "Modal Satisfaction"?

Answer: A LOT.

This post is arguing for the following account of modal satisfaction to replace McGinn's Tarskian one, as follows:

In any given modal statement:

1) only one predicate may be analyzed at a time.

The predicate being analyzed [i.e. being assigned a contingent/necessary operator] must be a predicate deploying the "is" of

predication, NOT the "is" of identity. This is because in order to evaluate any metaphysical property, we must keep all other predicates

constant, otherwise they cannot be legitimately compared, just as in a scientific experiment. Why? Because if we apply modal operators

to more than one given predicate at any given time then we cannot be sure whether one predicate s dependent on another.

E.g. Consider this statement:

"Socrates is a wise Athenean man".

We have the following predicates:

(i) "is wise".
(ii) "is an Athenean".
(iii) "is a man".

Now, to be an Athenean one must also be a human: a goat or a fish could not be an Athenean Citizens.

Women were excluded from becoming citizens (with limited exception in the later Hellenistic period).

Under Pericles' rule [461BC to 429BC], for example, in order for a marriage to be legal and the resulting children legitimate [i.e. legally

recognized as citizens of Athens], both the mother and father had to be Athenian citizens, and in addition the father and mother of the

married couple also on both sides, had to be Athenian citizens.

Hereditary links however, did not just determine citizenship. From the time of their birth, young Athenian men were expected to attain an

education. In Hellenistic Greece, a potential citizen spent 2 years in the gymnasium, and 2 years training in the military [also known as the

ephebeia]. Only after they had completed this education could the men then be deemed Athenean citizens.

So, in orderfor Socrates to be an Athenean Citizen, he needed to have done the following:

a) Been sired by an Athenean family [i.e. a family who were themselves already Athenean Citizens].
b) Attended a Gymnasium and the Ephebeia for their education.
c) Been a man.
d)Been a human being.
e) Been a mammal.

I.e. To be a citizen Socrates needed to have had the prerequisite education at a gymnasium and an Ephebeia.

But he could only have been able to attend a Gymnasium and an Ephebeia if he was a male.

But he could only have been a male if he were human.

But he could only have been human if he were a mammal.

etc, etc.


My pedantic and long-winded point here is that each predicate is parasitically ontologically dependent on something else

holding/being true.

For a single predicate assigned to Socrates, "being an Athenean Citizen", we see that there are a myriad of other predicates/facts

that must be true for this to be the case.

One cannot simply change one predicate without being fully aware of what we are implicitly entailing/implying.

To do so one would be guilty of metaphysical-modal sloppiness of the highest order.


So, to return to the statement:

"Socrates is a wise Athenean Citizen",
we see now that we only modally change one predicate if that predicate is not modally

dependent on another.

We must therefore only modally-modify one predicate at a time, and keep all others constant, otherwise we cannot identify

where the modal change has taken place.

If we modally-modify a predication deploying the "is" of identity", rather than the "is" of predication, then we are changing the nature of

the subject. And how can one analyze a subject/entity if we have changed its very nature!

I argue that we cannot.

Therefore, we can only modally example "is"'s of predication when dealing with a specific subject, one at a time.

N.B. I am not claiming that we cannot modally scrutinize more than one predicate at a time: I am merely stating that we cannot

modally-scrutinize a second [or third, or fourth] modal predicate until we have sufficiently analyzed it at the first level.

I.e. After one has analyzed the modal implications of all four of the original modal permutations of the statement "Socrates

is a man and is wise", THEN we can scrutinize both these elements together, but not before!


9:44 PM  
Anonymous JL said...

OK. Well, there's a lot of stuff to think about there. I agree that in a way we need to modal modify the "is" of predication, as in, modalising the copula. I didn't have room here to go into details concerning my own view of what the copula is (a function from complete to syncategorematic expressions) and how I think it is modified by modals.

I would allow that this copula function is multigrade, so that as well as being able to take single values like "handsome" and yield the predicative expression " handsome", they can take other numbers of predicate words, e.g. "tall", "dark" and "handsome". and yield the expression " tall, dark and handsome". The modal modification then only acts upon the copula function. Perhaps, for example, allowing us to form the expression "...could have been tall dark and handsome", which may then be satisfied or saturated by an individual, e.g. "Ronnie Corbett could have been tall, dark and handsome".

Whether this is true depends upon whether Ronnie Corbett could indeed have been tall dark and handsome. At this point, I only have a story about the logical form of how modal statements might be built up, which doesn't treat them in a radically different way to non-modal statements. As yet, I can't say anything more about what conditions in the world must hold for such modal statements to be true (or false).

2:49 PM  
Anonymous Anonymous said...

"As yet, I can't say anything more about what conditions in the world must hold for such modal statements to be true (or false)."

Mmm, the gist of my previous long-winded post was that we cannot manipulate modal predicates, or more than one given modal predicate at a time, unless we already know what other predicates in the sentence are modally dependent on it.

My example of the initially seemingly innocuous modal statement concerning Athenean Citizenship [laboriously] demonstrated this.

In effect, my conditions on modal satisfaction require a form of "logical omniscience".

This is because, being finite beings, it is difficult for us to foresee all ontological and modal relations between statements/propositions and their constituent predicates respectively.

Consider your example of Ronny Corbett:

Now, as discussed in my previous post, one needs to keep somethings constant while we vary the variables.

We need something fixed [i.e the subject] while we investigate other properties from a modal perspective.

But the problem here is that this account presupposes an identity for the subject in question.

E.g. "Ronny Corbett is short and wears glasses".

It is surely contingent that he is short. He may have been nutritionally deficient as a child, or perhaps his hereditary genes could have been otherwise.

It is contingent that he is/was a comedian. And it is modally possible that he is married to Angelina Jolie.

But the problem with any such hypothetical modal statement is that each such property, e.g. being married to Angelina Jolie [in a very distant possible world] or being tall, entail other things taking place.

E.g. To modally assert that it is possible that Ronny Corbett is tall, we are implicitly stating that it is possible that his parents carried genes which caused accelerated growth, or that his parents made him drink more milk as a child.

But then let us imagine that Ronny Corbett was brought up in a region of the world where there were very little dairy products available, e.g. China.

So, to assert that it is possible that Ronny Corbett could have been tall and that he drank a lot of milk as a child, one is also implicitly saying something about the availability of milk in China in the 1950's/60's [i.e. when Ronny was a child].

For milk products to readily available in China this would mean that there would have to be cattle available to be milked in China in the 1950's and 60's.

And if there wasn't any cattle readily available in China in the 1950's and 60's, then this would necessitate that China would have to acquire its dairy products eleswhere, and in the 1950's the worlds biggest trading country was, surprise, surprise, the United States of America [immediately after WWII, Europes economic power was smashed].

But for China to trade with the U.S, this would require that China and the U.S had already opened up diplomatic channels and were trading with one another.

Furthermore, as dairy products "go off" rather quickly, this would necessitate a good preservation method for the milk to be transported from the U.S to China.

So this would require the U.S to have developed some moderately advanced preservation techniques for [fresh] dairy products, or the use of mobile refrigerating systems to transport the milk.

So, from the harmless statement "It is possible that Ronny Corbett was tall because he drank lots of milk as a child" could implicitly require many other things to be possible or necessary.

The above argument seems to indicate to me that modal logic is nonsense on stilts: too confused and not formulated properly: just another case of philosophers tagging on another symbol and making it stand for a word.

It is hard to see what a diamond or square can do that the words "possible" and "necessary" cannot.... (;-)



7:12 PM  
Anonymous Anonymous said...

Consider another extension of the modal assertion that "It is possible that Ronnie Corbett is tall".

We already discussed the possibility of Ronnie Corbett growing to be tall due to consuming lots of milk as a child.

But what about the modal possibility of Ronnie growing up to be a tall man simply due to his genetic make-up? I.e. if his father carried a gene [or more likely, genes {plural}] which he hereditarily passed onto his son Ronnie...?

So the modal statement "It is possible that Ronnie is tall" would be replaced by "It is possible that Ronnie's father carried a gene [or genes] that caused Ronnie to grow taller".

But let us imagine that, in the 'real' world, Ronnie's father has a medical condition resulting in a growth hormone deficiency. This is usually genetically determined, so if Ronnie's father had it, it is highly likely that his father had it, and his father's-father had it,etc,etc.

So for the modal claim "It is possible that Ronnie Corbett was tall", to be modally true, we are also implicitly stating that it is possible for Ronnie's father, grndfather, great grandfather,etc,to have also been tall.

i.e. Facts and states of affairs are no atomic and isolated: they are interconnected with a myriad of other facts and states of affairs. So when stating a modal possibility, we must take care not be led astray by language, where one can simply negate one predicate by placing a square or a diamond in front of it....


1:11 AM  
Blogger bloggin the Question said...

I thin all R is trying to say is that modal sentences do not have independent truth values, but are rather dependent on sets of presuppositions, what I am calling epistemic frames of reference. Possible world semantics can deal with this, but can modalising the copula?
What I feel is appealing about the idea is that when talking about possibilities the tense is important, and the tense (in english) is often given by the copula. Sadly my grasp of grammatical theory is weak, but here are a list of statements concerning Ronnie Corbet (RC) and the predicate conjunction Tall Dark and Handsome (TDH)
1. RC is TDH........False
2. RC was TDH .......False
3. RC will become TDH ......unknown, not likely
4. RC might be TDH .......False (because 1 is false)
5. RC might have been TDH..... False because 2 is false.
6. RC might become ..... true because 3 is unknown.
7. RC could be TDH ..... No independent truth value. Needs to be completed with frame of reference. (Eg if given brain transplant into Richard Samuel's body)
8. RC could have been TDH ...... As above.
9. RC could become TDH .... as above but is verifiable to a certain extent since if RC does become TDH then 9 becomes retrospectively true.

What is interesting me is that one might try and eliminate tense by introducing a time index. But the copula is still there connecting the truth predicate to the time indexed proposition.
10,11 and 12. "RC is TDH" is false at 2007 and before but is unknown forward from 2007.
"RC is TDH" can not have been true from 2007 backwards but might be true forward from 2007.
"RC is TDH" could be true at any time given the right counterfactual assumptions.

My point here is that you cannot eliminate the effect of the copula by putting in a time reference since the copula just moves out of the proposition into the connection between the truth predicate and the proposition.

The general point is that it is not a good idea to think of propositions as timeless items constituted by their truth conditions in a static logically determined multiverse of possible worlds. It is better to think of propositions as belief states of free agents in the flux of time who use beliefs to guide actions freely through the world. What is, was will be might have been might become could have been and could be are very importantly different in this case. The world ios not a static collection of fact, but a sea of objects with ever changing properties in which we actively participate. So modalising the copula is the way forward.

RC could

3:20 PM  
Blogger bloggin the Question said...

Eliminating the copula:
Can the copula be eliminated using time indexes and world indexes? The idea is that they can all be reduced to "is" (and perhaps "are" for plural) with world and time index.
So "Ronnie corbet might have been tall" is reduced to "Ronnie Corbet is tall at world w and time t".
My argument against the reduction is that we can still ask whether this proposition is true. Now we have another copula. (RC is TDH, w, t) is true, or maybe it isn't true. Or perhaps it was true, or could have been true, or might be true. The copula elimination has failed.

6:56 PM  
Anonymous Anonymous said...

Interesting thoughts.

Sadly, on the time index front we have ourselves a case of what I like to refer to as "Anticipatory Plagiarism": where someone takes your ideas, steals them, and publishes them before you were born. (;-)

Arthur Prior concocted a tense logic in the late 50's/60's.

In addition to the standard logical operators, Prior added the following four "tense" operators:

P =

"It has at some time been the case that …"

"It will at some time be the case that …"

"It has always been the case that …"

"It will always be the case that …"

where "P" and "F" are 'weak' tense operators for obvious reasons.

Then there's 'Binary temporal Operators', S and U ("since" and "until"), introduced by Kamp (1968).

Where 'S' and 'U' mean:

"q has been true since a time when p was true"

"q will be true until a time when p is true"

Then there are lots of other tense operators, such as the "Metric Tense Operator", also by Prior, e.g. Fnp means "It will be the case the interval n hence that p".

And the "Next-Time Operator", 'O', where each subsequent statement/predicate is regarded as having occurred an incremental [positive] value to represent discrete time succession.

I highly recommend "Tense Logic "(1947) by Hans Reichenbach.

I agree with Jonny that the tensed copula is a problem.


1:40 AM  
Anonymous Anonymous said...

Another main point/problem to note concerns my initial post, where I mentioned that Boolos (1993) argued that modal logic has proved useful for central results concerning provability in the foundations of mathematics.

E.g. Existence proofs: the mathematician wants to determine whether or not a mathematical a solution exists [i.e is possible] for the problem, or whether or not it is possible that there exists a unique solution, and whether this solution is necessarily the only solution value, etc.

Modal logic has indeed proved extremely fruitful for the mathematician and mathematical logician.

But the problem here is that, as is well known, mathematics and mathematical logic operate in a very different way to normal statements for the simple reason that in a mathematical statement, the negation of one statement entails its contrary.

E.g. if the mathematician wants to find out if there exists a unique solution for a problem, he can find a second solution to demonstrate that there does not exist a unique solution.

But in philosophy, when negating a statement, or a predicate, it is hard to see in what sense the statement has a contrary [negational] equivalence.

What does this have to do with modal logic?

Well, consider the statement:

"Socrates is a wise mortal".

and the modal statement:

"Socrates is necessarily wise and necessarily mortal".

If the metaphysician wishes to investigate the metaphysical status of this statement using modal logic, he can experiment with the modal operators and negations:

"Socrates is necessarily wise and necessarily mortal".

But what if we negate parts of this statement?

We can then derive the following statements:

1) "Socrates is NOT necessarily wise and NOT necessarily mortal".

If something is not necessary, this doesn't entail that it isn't possible.

So [1] above is equivalent to:

"Socrates is possibly wise and possibly mortal".

But from a metaphysical perspective, it is hard to see in what sense, if any, statement [2] is a 'proper' negation of statement [1].

Or consider negating statement [2]:

"Socrates is NOT possibly wise and NOT possibly mortal".

If something is not possible, then it is impossible, and, ipso facto, also (trivially) not necessary.

It seems then, that we have three modal values: necessary, possible, and impossible.

So a double negation does not leave the truth value invariant, as in classical logic.

It is difficult to see the ontological status of such statements, especially when we're negating statements where, unlike mathematical statements, a negation does not entail its contrary, as there isn't necessarily a contrary to an ordinary language statement. And when we add modal predicates to ordinary language, everything gets messy...


12:32 PM  
Blogger bloggin the Question said...

Here's a modal problem that was in Jessica's original post in terms of instantiating, but also relates to this last points about negation.
Let us suppose that we don't know whether or not the bun is in the oven.
1. The bun might be in the oven....true.
Modal inference: Possibly p entails possibly not p. If it might be that P then it might be that ~p. If the bun might be in the oven then the bun might not be in the oven.
2. The bun might not be in the oven.......true.
2nd Modal inference: Necessarily p entails not possibly not p. If the bun is not in the oven then it is not true that the bun might be in the oven.
Likewise: If the bun is in the oven then it is not true that the bun might not be in the oven.
LEM. Pv~p. Either the bun is in the oven or the bun is not in the oven.
3. In which case either the bun might be in the oven is false or the bun might not be in the oven is false.
Statement 1 entails 2 but the conjunction of 1 and 2 contradicts 3.

12:56 PM  
Anonymous Anonymous said...

Yes, this is the sort of thing I have in mind.

What Jessica's highly interesting case of the bun in the oven indicates, so me at least, is a form of reductio ad adsurdum against Modality without caution.

To return to Jonny's earlier comment about indexical time:

I agree that this is an acute problem, and my proposed solution, as with many philosophical "solution", tends to raise more problems.

Jonny wrote:

"The general point is that it is not a good idea to think of propositions as timeless items constituted by their truth conditions in a static logically determined multiverse of possible worlds."

Very true.

"It is better to think of propositions as belief states of free agents in the flux of time who use beliefs to guide actions freely through the world."

Mmm, but this would render modal logic trivial, as the truth value of a belief state is true iff it is thought of by the person. E.g. Jimmy believes that in another world, it is possible that Cafe Nero sell chocolate coated panda meat.

It it trivially true that Jimmy believes this, and is therefore always true, that Jimmy believes it.

"Can the copula be eliminated using time indexes and world indexes? The idea is that they can all be reduced to "is" (and perhaps "are" for plural) with world and time index."

I'm reminded of a line from T.S. Elliot's "Four Quartets" [I think it was that poem, though I'll have to check].

The line is "To be conscious is to be outside time".

This always puzzled me as a child, but I feel it to be relevant to the current discussion:

To be conscious is to be outside time, i.e. when we reason we use language.

Think of all the recent discussions you have had with others, face to face, on the phone etc. People normally refer to concepts, ideas, where you're going to go, what you're going to do, what you've done, how your day was, etc, etc. We rarely ever actually refer to the here and now in our everyday conversations.

This is bizarre, and a very interesting psychological observation.

All our modal statements refer to states of affairs.

Jonny is right that we cannot refer to statements as being timeless and discuss them and their modality without reference to the here and now.

So how could we bypass this?

I would argue that we solve this by never referring to the present in modal statements.

Instead, we always refer to either the past or the future.

The temporal now is constantly fleeting, so in a sense, whenever we refer to something it will have already taken place.

The only metphysically solid frame of reference is the past.

We can then engage in idle philosophical speculation with modality pertaining to past events.

This is a common device in fiction, the "what if" ideas. E.g. Philip. K. Dick's "The Man in the High Castle", among many others.

We can, of course, also refer to the future. Again, another common device of science fiction writers [all of them infact].

But the present should therefore be off limits to the modal logician.

The reason we can manipulate with the past using modal operators is that, as we already know the 'true' state of affairs, we therefore already have a state of reference: history!

We can therefore manipulate modal predicates pertaining to past events.

Anything else is idle science fiction, and to engage in it we would cease to be philosophers, and become science fiction writers....

But then again, is this such a bad thing? Arthur. C. Clarke went to King's, and surely a science fiction novel would pay better than a dry academic textbook on modal logic... (;-)



9:19 PM  
Anonymous Johnny Austin said...

At least you write in paragraphs, unlike the usual poster to this blog. "What can be said at all, can be said clearly" is certainly true in your case

9:49 PM  
Anonymous Jonny said...

- "What can be said at all, can be said clearly" is certainly true in your case -
Surely if the nested sentence is true at all it is true in every case. That it is certainly true in Jessica's case seems to imply that it may not be true in somebody else's case. This would make the following possible:
Jessica can only say what can be said clearly, whereas Johnny Austin can say things that can't be said clearly.
If this were true then there would be a class of things that Johnny Austin was capable of saying but that Jessica Leech was incapable of saying.
In this light to say that "What can be said at all can be said clearly" is true in a particular person's case is to attribute an incapacity to that person, an incapacity that is not generally shared.
I hope that it isn't true in my case. I hope that I can say things that can't be said clearly. It sounds like a sexy creative thing to be able to do.

4:08 PM  
Anonymous Anonymous said...

Jonny Austin's right though.

Most, if not all, the individuals posting in this blog are either MPhil or PhD students. And as such, our prose should be polished and succinct, minus the spiele of many undergrad [and continental, *cough*] writing.

As such, I always try to adhere to the following rules:

1) Don't try to be clever: just try to be understood. As such, there's no need for fluffy erudition that may muddy the waters of thought.

2) Don't insult people. This can be summed up with the following slogan: "Attack propositions, not people!".

Admittedly, I may miss the mark occasionally...


Re. Jonny's comment:

"- "What can be said at all, can be said clearly" is certainly true in your case -

This is clearly true.

If you can't explain something to someone [of moderate intelligence] then the odds are that you don't fully understand it yourself.

As Einstein aptly put it:

"Make things as simple as possible; but not simpler..."

It is trivially true that all statements and descriptions are underdetermined by language.

I.e. Any given statement/proposition can be phrased and paraphrased in a multiplicity of ways: some of these phrasings may be clear, succinct and lucid. whereas others may be muddled and confused, despite the fact that they both aim to express/state the same point.

On this point Johnny Austin is on the mark.

Jonny then writes:

"Surely if the nested sentence is true at all it is true in every case. "

Of course, this again is due to the fact that as any given statement can be phrased in a multiplicity of ways, some ways will be more clear than others.

But the generalization does not follow, and this is due to the following reason:

Some propositions are simpler than others.

Consider a European Directive or any other Law: it will usually contain a vast number of clauses and be rather long and tedious.

Now,these laws appear to me, someone with non-legal training, to be extremely unclear, but they are as clear as they could be.

Consider the Income Tax Act, 2007:

Laws must be as explicit as possible in order to avoid any loopholes of interpretation in permitting a judge to set a precedent in interpretation for an individual case.

As such, Jonny's statement:

"Surely if the nested sentence is true at all it is true in every case. " is rendered false for generalizations.

Jonny continues:

"Jessica can only say what can be said clearly,".

Referring back to my earlier comment [in this post] concerning the underdetermination of language, this is false, because Jessica, using the rich linguistic tools at her disposal, is free to use whichever words to form whichever phrases she so wishes.

She is therefore free to speak clearly or unclearly.

For her to always speak clearly, it therefore follows that either a) Ms Leech can only speak clearly because she lacks free will, or b) Ms Leech lacks the vocabulary in which to speak 'un-clearly'.

It is important to note that the very notion of speaking 'clearly or 'unclearly' is a relative term.


Because some people will understand what she's talking about while others may not.

E.g. Lots of academic disciplines deploy their own jargon. These jargon words may appear to be gibberish to a layman, but actually serve to facilitate communication between their fellow peers.

The same applies to sub-cultures: different cultural groups use their own slang: when conversing with other members of society who share their culture, these slang words will be meaningful and appear clear to them, but to an outside they will be deemed unclear.

As such, the dichotomy between 'clear' and 'unclear' is a vague one of heuristic value only, and with no real tangible utility within philosophy as an academic discipline due to being a relative term.


Is it just me, or do philosophers have a capacity of overcomplicating things...? (;-)

10:58 PM  
Anonymous Anonymous said...

The law link didn't come out properly, so here's a shorter link containing links to many laws.

If you read through one of two of them then you should see what I mean Re. unclearness being relative.


11:01 PM  
Anonymous Anonymous said...

If the website link in the previous post contained too many links to choose from then try this one:



11:06 PM  
Anonymous Anonymous said...


Jonny wrote:

"In this light to say that "What can be said at all can be said clearly" is true in a particular person's case is to attribute an incapacity to that person, an incapacity that is not generally shared.".

Mr Austin is correct here because his earlier statement, viz., that "What can be said at all can be said clearly" was a general statement referring merely to the logical linguistic possibility of forming a statement or proposition using language.

The linguistic possiblity is clearly true: given the rich resources of the English language at our disposal, it is usually possible to say it clearly.

The crucial point made by Mr Austin but missed by Jonny was the word "can", reflecting merely that it is 'possible' to state something clearly, not that one 'must' do so.

Right. I've now used up my pedantic quota for this blog now. ;-)


12:32 AM  
Anonymous Jonny Blamey said...

I can't really understand the point that R is trying to make. Here are some statements I take R to be making:
1. The following statement made by Mr Ausitn was true in context : "What can be said at all can be said clearly" is certainly true in your case. Let us name this statement PA=PC T(JL).
Let's accept for argument that "your" refers to Ms Leech. (It could have refered to R).
What I called the "nested sentence" is signified by PA=PC. It is a conditional statement which comes out as "if P can be asserted then P can be asserted clearly."
2. R asserts that PA=PC is a general statement that is not true in all cases. By "cases" R seems to mean something like statements, since R uses complex statements of law as examples of statements that can perhaps not be stated any more clearly, though they are not generally clear.
3. R asserts that "Clearly" is a relative term, since what is "clear" to one person maybe "unclear" to another. R makes what I take to be a false statement that relative terms have no place in philosophy, but I leave this to one side. I would like to point out that the type of relativity here is assessor relativity. This is undoubtably true.
4. R makes some general comments about the desirability of clarity which amounts to a defence of PA=PC.
The conclusion that I take R to be making is that Mr Austin's statement was true in context and is to be interpreted as meaning that if Ms Leech is permitted to assert that P, then she is obliged to assert that P clearly.
However I don't think that R's comments support this conclusion.
Firstly, while it is true that PA = PC is a general statement so it is possible that it is not true in all cases, "cases" seem to be statements not people, so to say that PA = PC is true in "your" case doesn't make sense, unless "your case" applies to a specific statement.
But perhaps PA = PC is relative because "clearly" is relative, in which case it could be true to some people, but not to others. This would make PA = PC T (JL) mean something like: if something can be said to J.L. at all, it can be said to her clearly. This seems to be just plain false, and not at all what Mr Austin meant.
Another possibility is that "can" is normative rather than signifying a capacity. So rather than meaning if J.L. is capable of saying something at all she is capable of saying it clearly it means that given she is a Research student in philosophy the norms dictate that if she is permitted to say something at all she is permitted to say it clearly. But this doesn't seem quite right because it is trivially true. Although "can" can be interpreted as "am permitted to", it is rarely interpreted as "am obliged" although it can occasionally be used in this way as in "You can bring me a present next time you visit me." This would render J.As statement as "If J.L. is obliged to say something, she is obliged to say it clearly, since she is a research student of philosophy."
Actually, this works doesn't it. Perhaps I do agree with R and Johnny Austin after all. If Johnny Austin had been a little more clear in the first instance.... Perhaps the dictum doesn't apply in his case!

1:06 PM  
Anonymous Anonymous said...


Austin merely said that "What can be said can be said clearly".

The way you interpreted him initially had me flummoxed when you derived a bizarre [highly impressive] Reductio by comparing it with Jessica's argument.

My legal example was not stating that relative terms have no place in philosophy, in toto, I was merely pointing out that such relative terms have no discernible utility until they have been formalized.

The Philosopher usually begins his reply to another with the standard question: "What do you mean by 'x'?", or "Define your terms".

This formalizes a term, so that, even in the case that you may not agree with your philosophical 'sparring' partner, you at least know what each other is referring to.

This brings me back to the term 'clear'.

My legal examples merely demonstrated that something may be said but may not be clear to all.

My further example of esoteric academic jargon and cultural slang further buttressed this.

It is interesting to note, by way of passing, the great Sociologist, Max Weber's, distinction between 'Implicit' and 'explicit' speech: implicit speech is usually conducted between people who are familiar with each other [though this by no means entails they are friends]. Explicit speech, on the other hand, is usually conducted between individuals unfamiliar with one another.

Now, explicit speech patterns will typically be rendered 'clear' to an eavesdropper,whereas implicit speech will not, as it will typically be embedded within a wider mode of linguistic idioms, slang, jargon and personally shared experiences.

Jonny writes:

"The conclusion that I take R to be making is that Mr Austin's statement was true in context and is to be interpreted as meaning that if Ms Leech is permitted to assert that P, then she is obliged to assert that P clearly."

This is not what I was suggesting. My previous post addressed this point in referring to the underdetermination of language: viz., that any given statement may be phrased in a myriad of different ways, and some phrasings will appear more clear than others.

But this in no way obliges Ms Leech to use those particular words any more than another choice.

That is precisely what I was asserting, and it seems to coincide with your own views Jonny.

To assert that one ought to always phrase something clearly is an aesthetic preference that, although many would undoubtedly support it [who would intentionally prefer muddled prose to lucid ones?], it is just that: an aesthetic preference, and nothing more...

You continue:

"But perhaps PA = PC is relative because "clearly" is relative, in which case it could be true to some people, but not to others. This would make PA = PC T (JL) mean something like: if something can be said to J.L. at all, it can be said to her clearly. This seems to be just plain false, and not at all what Mr Austin meant."

This is not what I was saying at all! My legal example demonstrated this by pointing that something may need to be said, but it may not necessarily be clear! I.e. Laws must be phrased in a certain way, but this may prevent them from being clear.

And even if this were what I was saying, to assert that "This seems to me to be plain false" is an odd thing for a philosopher to say without stating why you find it false. But this is a different issue.

"Research student in philosophy the norms dictate that if she is permitted to say something at all she is permitted to say it clearly. But this doesn't seem quite right because it is trivially true."

Yes! I've already said this in an earlier post!

I explicited stated that it is trivially true that all statements are underdetermined in that any given statement may be phrased and rephrased in a multiplicity of ways!

I wonder if you've read my previous posts... or perhaps I wasn't clear enough.... (;-)


P.s. Now the term's over does this mean no more seminars = no more blog posts until September/October...?

10:14 AM  
Anonymous Anonymous said...

To avoid any further confusion I should formalize my views on clearness, which seems to have temporarily become the focus of this post on modality [how did this happen?].

clarity is a relative term because some people may find a statement clear whereas others may not.

This is due to the fact that for something to be deemed 'clear' is for it to be 'understood'.

However, understanding of a statement depends on:

1) Background knowledge of the topic being discussed.
2) shared used of language between the speaker and the listener [or the writer and the reader].
3) A sufficient cognitive ability.

Stated formally:

P1) For a statement to be judged as clear is for it to be understood.
P2) But for something to be understood depends on background knowledge relating to the topics/issues being discussed, a shared mode of speech [e.g. jargon words, the same language {obviously}].
P3) But different people possess different levels of knowledge and understand different languages.
Conclusion: the notion of 'clear' is a relative term.

We should make the distinction here between an 'absolute' term and a 'relative' term.

E.g. Black, white or flat are absolute terms: something is black absolutely [if you use language correctly], as it refers to a total absence of colour. White is an absolute term as it refers to all the colours in the spectrum. And flat is an absolute term.

Whereas the terms curved are relative because things can possess a different degree of curvature [anyone who has studied calculus or been on a rollercoaster will know this].

The same applies to 'clear': it cannot be posited absolutely due to it being a relative term.

Instead, it must be posited of the individual who utters it.

I.e. "That seems clear to ME".

"Do YOU find that clear?".



Now we have established the relativity of the term 'clear' we may pedantically progress:

The underdetermination of language entails that any given statement may be phrased in a multiplicity of ways. It is trivially true that one phrasing may appear more 'clear' to more people.

But this is in no way entails that it must be possible to make it clear.


10:31 AM  
Anonymous jonny blamey said...

I hope you're enjoying the bizarre humour of this thread as much as I am, R. Clarity is an interesting topic for sure, especially since it is itself a philosophical standard, as well as being like you say relative and also, as I am about to say metaphorical.
Interested as I am in probability and certainty I'm interested in the idea of seeing something approaching from a distance until some definitive moment when it becomes "clear" what it is you are looking at. Let us say a person. At a hundred yards, you may not be able to tell whether it is a man or woman or even a person at all. The range of possibilities will be broad and shallow. Think of it like a partition. There could be a thousand possibilities at a hundred yards, but at the distance of maximum clarity this could wittle down to one. Normal conceptions of probability will distribute probabilities adding up to one across the partition. As it becomes more clear what it is you are looking at the logical space will tend towards one possibility with a probability of one.

We can make this less metaphorical by imagining the object and observer to be staionary but for the medium between them to begin opaque and gradually become transparent. Keep the object as the facee of someone familiar to the observer and we may find that the point of complete recognition, in otherwords the point at which the observer has no doubts as to who the object is and will reasonably assign a probability of one to the statement that the object is Fred (or whoever it happens to be) comes at a point before the medium has become completely transparent.

Much more difficulty is applying this concept metaphorically to clarity of expression. Here the observer is the reader, or radical interpreter of a statement, the object is the utterance, and the medium is language. We can imagine the language used in the utterance become more and more clear until the reader finally can only interpret it as meaning one thing.
The question then is what is this "thing"? Answers are a set of truth conditions, a set of possible worlds, a proposition or a statement. However, I thinkk this is a dangerous kind of analytic tradition slothfulness and complacence, since the same analogy could be applied to an order, or a question.
The general picture is that as an utterance becomes clearer, the number of possible interpretations goes down and the probability of the remaining possible interpretations goes up until both reach 1 and understanding is perfect.
I'm trying my hardest to use paragraphs.

There is no reason to stop posting because of the summer, I'm very busy right now, but I intend to put up some posts in the near future, one by someone from LSE, one from Oxford and one about the switching paradox for subjective probability, which is giving me nightmares. If you don't mind breaking your anonymity, R, I wouldn't mind posting something that you send me.

5:19 PM  
Anonymous Anonymous said...


I am indeed enjoying the discussion: Clarity/Clearness is an interesting topic in itself.

Your discussion of the man in the horizon slowly becoming clearer with reference to probabilities is fascinating. I don't know if you're familiar with Bayesianism, but you're impicitly referring to that: you have your prior probabilities,and as more information gets acquired [viz., in this case, the increase information is acquired via moving closer towards the chap in the horizon], you sequentially update your Bayesian Probabilities to calculateyour posterior probabilities.

1.0 is certainty and 0.0 is impossibility.

I'd be happy to post something on the philosophy of probability actually, with reference to Bayesianism and Keynes and Ramsey's views, especially Keynes's, as expounded in his "Treatise on Probability Theory".

You can email me at



7:11 PM  
Anonymous Anonymous said...

btw, the extra 'l' is deliberate, as someone else as had already taken '' and ''.


7:13 PM  

Post a Comment

<< Home