A challenge to Frege on assertion

Pierre-Normand

On the other hand, in the wake of the Tractatus and Carnap and the rest, we got possible-world semantics; so you could plausibly say that a picture showing how things could be is a picture showing how things are in some possible world, this one or another.

The feeling of "claiming" is gone, but was probably mistaken anyway. In exchange, we get a version of "truth" attached to every proposition, every picture. Under such a framework, this is just how all propositions work, they say how things are somewhere, if not here. Wittgenstein's point could be adjusted: a picture does need to tell you it's true somewhere -- it is -- but it can't tell you if it's true here or somewhere else. — Srap Tasmaner

I was a bit puzzled by this so I asked ChatGPT o1-preview the following question:

Reveal

"Hi ChatGPT,

There is an ongoing discussion thread currently on ThePhilosophyForum titled "A challenge to Frege on assertion". It's inspired by a discussion in Irad Kimhi' recent book "Thinking and Being". I haven't read Kimhi's book yet but I have some acquaintance with background material from Frege, John McDowell and Sebastian Rödl. I also have some acquaintance with Wittgenstein's late philosophy but very little with the Tractatus.

In response to user Leontiskos who asked "...is it possible to see that something is true before going on to assert it?", user Srap Tasmaner replied thus:

'''
There's a couple ways to read this, but at any rate, a couple obvious options:

[snip]

[...] but it can't tell you if it's true here or somewhere else.
'''

I may be off the mark but the way in which possible world semantics is being evoked here reminds me more of Lewis than Kripke. If Srap Tasmaner is right in his interpretation of Wittgentstein's Tractarian picture theory of meaning, it would seem like what propositions (or "pictures") express (or "show") lack Fregean singular senses (rigid designators). On Kipke's picture, as well as McDowell's and Wiggins's understanding of Frege regarding singular senses, propositions about particulars (that are picked up by singular — non-descriptive — senses) can't be true of different objects (i.e. counterparts) in non-actual possible worlds. That's because they function as rigid designators. Setting aside for one moment the question whether or not McDowell's and Wiggins's understanding of singular senses aligns with Frege's own thinking, do you happen to know if Wittgenstein's Tractarian view of the picturing relations between language (or thought) and the world comport more with a "rigid" view of singular senses or rather with a more Lewisian "counterpart" view that would portray the "singular" components of pictures in covertly descriptive terms?"

In submitting my question to ChatGPT, I first mistakenly selected the model GPT-4o rather than the new ChatGPT o1-preview. The main difference between the two models is that the second one "thinks" before answering. Here is GPT-4o's response:

Reveal

It’s great that you’ve drawn attention to the tension between Wittgenstein’s picture theory in the *Tractatus* and the rigid designator framework as developed by Kripke, and possibly McDowell and Wiggins.

In the *Tractatus*, Wittgenstein's picture theory suggests that propositions function as "pictures" of possible states of affairs. These pictures are made up of elements that correspond to objects in the world, and their arrangement represents the way those objects could be related. The truth or falsity of the picture depends on whether the represented arrangement of objects corresponds to the actual arrangement in the world.

Now, regarding your question of whether Wittgenstein’s view aligns more with a Kripkean notion of rigid designation or with a Lewisian "counterpart" view, I would lean towards the latter — but only cautiously.

1. **Fregean/Kripkean view**: Frege's notion of a singular sense (or what Kripke later develops into a "rigid designator") implies that certain terms in a proposition refer directly to specific individuals, across all possible worlds. Kripke's "rigid designator" is closely tied to the identity of the object itself, rather than any descriptive characteristics it may possess in those worlds. For example, "Aristotle" rigidly refers to the historical philosopher Aristotle, no matter what properties he has in different possible worlds.

2. **Wittgenstein's picture theory**: Wittgenstein, in the *Tractatus*, doesn't directly engage with the notion of rigid designation or counterpart theory, but his treatment of propositions as pictures suggests that the components of a picture (which correspond to objects) can only represent a possible state of affairs without making claims about necessity across all worlds. The elements of the picture aren't inherently rigid in their reference; rather, their reference depends on how the picture relates to a possible arrangement of things. This is why some commentators see Wittgenstein’s view as more closely aligned with a descriptive theory of reference. The picture only tells us what could be the case, not necessarily what *is* the case about a particular rigid object.

This would seem to align with a **Lewisian "counterpart" theory** rather than Kripkean rigid designation. In counterpart theory, individuals in different possible worlds are related by resemblance, not by identity. So, Wittgenstein’s picture theory might suggest that the elements of a picture could correspond to different "counterparts" in different worlds, without the insistence on rigid identity. The “picture” doesn't necessitate that the same individual exists across all possible worlds, which contrasts with the Kripkean idea that a rigid designator must refer to the same object in all worlds where it exists.

3. **What about the singular sense in Wittgenstein’s view?** The key difference is that Wittgenstein doesn’t seem to be as concerned with the rigid reference of singular terms in the way that Frege, Kripke, or even McDowell and Wiggins are. The *Tractatus*’s emphasis is more on the structural relationship between the elements of a proposition and the world rather than on a theory of singular reference. So, his theory seems neutral or even in tension with rigid designation, given that a picture simply shows a possible arrangement and doesn’t guarantee any trans-world identity for its elements.

To summarize: while Wittgenstein's *Tractatus* isn’t directly engaging with possible world semantics as later philosophers like Kripke or Lewis do, his picture theory appears more amenable to a Lewisian "counterpart" view, where the elements of a proposition could be covertly descriptive and differ across possible worlds, rather than Kripke’s rigid designation where they would have to refer to the same entity across all possible worlds. However, since Wittgenstein wasn’t working with these concepts in the same way, it’s more an issue of interpretation and alignment rather than a definitive stance on rigid designation versus counterpart theory.

And here is ChatGPT o1-preview's response, including its own summary of its initial thinking episode. I checked the accuracy of its quotations from the Tractatus and it got them right except for 2.022 that actually reads "It is obvious that an imagined world, however different it may be from the real one, must have something—a form—in common with it." The content o1-previews quoted as 2.022 is actually very close to 2.0271, which reads: "Objects are what is unalterable and subsistent; their
configuration is what is changing and unstable."

Reveal

Thought for 12 seconds

-- Beginning of ChatGPT's internal thinking --

Clarifying Wittgenstein's perspective

The user examines Wittgenstein's Tractarian view, aligning it with singular sense rigidity à la Kripke, or a Lewisian counterpart theory allowing counterparts in different possible worlds.

Clarifying picture theory

I'm exploring Wittgenstein's picture theory, examining whether its components align with Kripke's rigid designators or Lewis's counterpart theory.

Sketching connections

Wittgenstein’s ideas about propositions and reality’s correspondence resemble Kripke’s rigid designators and Lewis’s counterpart theory, which contrast how objects and names function.

Evaluating Wittgenstein's names

I'm thinking through whether Wittgenstein's names act as rigid designators, universally identifying the same object, or as descriptions that could denote different objects across various possible worlds. This involves analyzing philosophical perspectives from Kripke, McDowell, and Wiggins on singular senses in Frege's framework.

Clarifying Wittgenstein's view

I’m piecing together Wittgenstein's notion that in the Tractatus, names consistently point to the same objects across different possible worlds, reflecting their stable reference to simple, persistent entities.

Analyzing language structure

OK, let's think it through: Wittgenstein's Tractatus uses rigid designators, consistently referring to the same objects across all possible worlds, differing distinctly from Lewisian counterpart theory.

-- End of ChatGPT's internal thinking --

In Wittgenstein's Tractatus Logico-Philosophicus, the picture theory of language posits that propositions are logical pictures of states of affairs. These pictures are composed of elements—names—that correspond directly to objects in the world. The structural relations among these names in a proposition mirror the relations among the objects in reality. This mirroring allows the proposition to "picture" a possible state of affairs.

Rigid Designators in the Tractatus

In the Tractatus, names function as direct references to objects without the mediation of descriptive content. Wittgenstein emphasizes that names are simple signs that stand for simple objects. For instance:

2.02: "Objects are simple."

3.203: "A name means an object. The object is its meaning."

This direct reference aligns closely with the notion of rigid designators introduced by Saul Kripke in his seminal work Naming and Necessity. A rigid designator refers to the same object in all possible worlds where that object exists. It does not pick out different entities in different possible worlds based on descriptive content; instead, it consistently refers to the same object across all possible scenarios.

Comparison with Fregean Singular Senses

Frege's concept of sense (Sinn) involves the mode of presentation of an object, which can be thought of as the way the object is given to the mind. In discussions by philosophers like John McDowell and David Wiggins, singular senses are understood as directly referring to particular objects without reliance on descriptive content. This interpretation of Frege aligns with the Kripkean notion of rigid designators.

Wittgenstein vs. Lewisian Counterpart Theory

David Lewis's counterpart theory suggests that individuals in different possible worlds are not the same entities but counterparts—distinct individuals that resemble each other in certain respects. Under this view, a term might refer to different objects in different possible worlds, based on descriptive characteristics.

In contrast, Wittgenstein's Tractatus does not support this counterpart approach. Since names in the Tractatus are directly linked to specific objects without descriptive mediation, they function more like rigid designators. The objects that names refer to are the same across all possible worlds (or possible states of affairs). Wittgenstein writes:

2.022: "The object is the fixed, the existent; the configuration is the changing, the variable."

2.027: "Objects, the unalterable, and the subsistent are one and the same."

These passages suggest that objects are constant and unchanging—they form the substance of the world. Therefore, names consistently refer to the same objects across all possible configurations of facts.

Conclusion

Given the above, Wittgenstein's Tractarian view of the picturing relationship between language (or thought) and the world comports more with a "rigid" view of singular senses. The elements of the picture (names) are directly tied to specific objects, functioning as rigid designators that refer to the same entities across all possible worlds. This stands in contrast to a Lewisian "counterpart" view, where terms might refer to different objects in different possible worlds based on descriptive similarities.

Therefore, in the context of your question, Wittgenstein's Tractatus aligns more closely with Kripke's rigid designators and the Fregean singular senses as interpreted by McDowell and Wiggins, rather than with a counterpart theory that treats singular terms in covertly descriptive terms.

Banno

↪J

,

↪Srap Tasmaner

As I understand it, 0th level terms - a,b,x - name individuals. Included amongst those individuals are propositions, which in turn name a truth value. So (2+2=4)=(5-3=2)= ⊤. These all name the same individual. So "Berlin" might name the city, and "2+2=5" might name the false.

Deciding what these terms refer to is done when one gives an interpretation to the letters.

Propositions have truth values in virtue of being names for the true or the false.

"Berlin" is not usually interpreted as a name for either true or false, so it doesn't have a truth value.

“What would Frege say about comprehending a singular term?” — J

On this account, both "Berlin" and "2+2=4" are names. Indeed, if a proposition is considered to be a statement with a truth value, then any proposition is just the name of either the true or the false. Assuming bivalency, of course.

Frege sets up existence in terms of satisfaction - though he doesn't use that term. He uses the concavity as a version of the universal quantifier. So far as existence is defined, it is defined in terms of universal quantification. It was working through this that led to Model Theory. So there is much to be said here.

frank

I'm happy to be shown otherwise. — Banno

He has a pretty rigorous argument for indefinability. I think J wanted to see it too.

Srap Tasmaner

I was a bit puzzled by this so I asked ChatGPT o1-preview the following question — Pierre-Normand

Just so you know, I'm not going to read that.

Banno

↪frank

Sure.

And every other attempt to define truth collapses too. For in a definition certain characteristics would have to be stated. And in application to any particular case the question would always arise whether it were true that the characteristics
were present. So one goes round in a circle. Consequently, it is probable that the content of the word " true " is unique and indefinable — The Thought: A Logical Enquiry

A bit later we get Tarski's definition of truth in terms of satisfaction, getting around the recursion by using metalanguages. I'm contending that what Tarski does is a working through of the implications of Frege's system; that satisfaction is to a large extent implicit in Frege.

Edit: Satisfaction is a pretty straight forward notion. "The cat is on the mat" is true if the cat is one of those things that satisfies "...on the mat" just means that "the cat is on the mat" is true if the cat is one of the items in the group of things "...on the mat".

Pierre-Normand

Just so you know, I'm not going to read that. — Srap Tasmaner

That's fine, but you can read merely my question to it since it's addressed to you too. I had thought maybe Witt's Tratarian picture theory of meaning could comport with a Lewisian "counterpart" view (since I'm not very familiar with the Tractatus, myself) but Frege's own conception of singular senses comports better with Kripke's analysis of proper names as rigid designators (which is also how McDowell and Wiggins also understand singular senses in general -e.g. the singular elements of thoughts that refer to perceived objects). The quotations ChatGPT o1-preview dug from the Tractatus seem to support the idea that this understanding of Fregean singular senses is closer to Wittgenstein's own early view (as opposed to Carnap's or Lewis's), also, although Witt eventually dispensed with the idea of unchangeable "simples" in his later, more pragmatist, philosophy of thought and language.

Srap Tasmaner

Is there then something between assertion and non-assertion? — Leontiskos

Almost everything I've posted in this thread.

There are a couple places where I was moved to think, but nothing much came of it.

Most of my posts have been pretty lazy, stuff I can post on auto-pilot. Sometimes I do this just in case I happen to know something someone else would find useful. If you asked me if I believe any of it, I'm not sure what I would say.

I like the stuff about the screwdriver. And I'm right about triangulation. Everything else was chit chat. Philosophical gossip.

frank

↪Banno

We disagree about the value of Tarski's work. You don't need Tarski to play around with the T-sentence rule, though. I think that's what you're doing.

Banno

↪frank

Ok. Whatever else you might think about truth, it's pretty hard to disagree with Tarski. Is that what you want to do?

frank

Ok. Whatever else you might think about truth, it's pretty hard to disagree with Tarski. Is that what you want to do? — Banno

You don't get Tarski unless you understand that his truth predicate doesn't mean anything. It's not the truth we talk about in ordinary language. Don't equate Tarski and the T-sentence rule. The latter has become a philosophical gadget in a variety of projects.

Banno

↪frank

Sure, it's a piece of mathematical logic. That doesn't render it meaningless. So are you taking on Tarski, the relevance of which to a discussion of Frege and mathematical logic is obvious, or perhaps Davidson, for daring to suggest that first-order logic might tell us something about language? Which?

frank

↪Banno

I think that like Davidson, you have reinterpreted Tarski for your own purposes. No shame in that.

Banno

↪frank

Ok. I haven't used any of Davidson's ideas here. So I can't see how that impacts this thread.

frank

↪Banno

This is the sentence in the OP that stood out to me:

An assertion can be displayed, perhaps as an integral part of a proposition, without being an “actual assertion. — J

I believe it follows from Frege's view of truth that assertion is integral to propositions. Think of the scrap of paper that says "Berlin.". Any effort to read it as a proposition will immediately conjure situations in which someone is asserting it. The mind is wired to look for purposeful communication.

When a logician separates assertion from proposition, meaning becomes unstable. The only source of stability is the potential for assertion.

Banno

↪frank

Ok. Not sure what that has to do with Davidson and Tarski.

"The cat is on the mat" is a string of letters. It's also a well-formed sentence in English, making it a locution. It can be used, amongst other things, to talk about the spacial relation between the cat and the mat. But it might also be used by a secret agent to give the order "Attack at dawn!". The step from a locution to an illocution requires some sort of interpretation, in which perhaps the parts are given a reference, perhaps some other sense. It might be used to say that it is tru that the cat is indeed on the mat, to assert that the state of affairs presented is indeed the case. It might also be used to ask if the cat is on the mat: "The cat is on the mat?". That it has the grammatical form of a statement does not imply that all that can be done with it is to make assertions. If it is used as something that can be either tru or false, then we have some grounds for calling it a proposition. We can consider the content of the sentence, the cat and the mat, and understand which cat and which mat, and yet not know if it is being used as a statement or a question or for something else entirely - perhaps to say that the dog, who sleeps on the mat whenever it is inside, is outside - The cat is on the mat" means that the dog is outside. And then there are perlocutions.

And so on.

All this is not definitive, nor complete, but just a list of some of the aspects that might be considered from even such a simple string of letters. It gets complicated, quickly.

Now one way of dealing with statements is to seperate out the things that are at least prima facie referred to, and to examine the structure of those references. As it turns out Frege showed how we might do this by placing the judgement stoke at the beginning of the expression, thereby separating out the extension from it's locutionary force - although he would not have described what he did in this way. Doing this allowed the development of a quite sophisticated area of investigation, that ultimately helped with developments in truth theory, Model theory, undecidability, modal logic, set theory and so on, and which is central to the workings of device on which you are reading this.

That looks to me to be quite tho opposite of "When a logician separates assertion from proposition, meaning becomes unstable".

So I don't understand your thinking here.

Forgive my extemporising.

frank

Forgive my extemporising. — Banno

:up:
In a way, the OP is asking about the extent to which meaning is use. In what circumstances can we drop use and still have meaning? This is assertoric force:

'assertoric force follows if a proposition expressed by an indicative clause is presented as relevant in a context made up of a subset of the hearer's factual assumptions."

You can already tell from that definition that assertion is generally going to be involved in the meaning of a proposition because you need to know context to discern what proposition is being expressed. Math is is different situation though. It has a semi-stipulated character, in part to avoid a paradox arising from Frege's work. This tells us that when we're talking about Frege, we're talking about a perspective that sees math as a matter of description, specifically of the contours of the human mind: of thinking. In the light of that, it becomes more interesting to consider his thoughts about separating sense and use.

More extemporizing.

On this account, both "Berlin" and "2+2=4" are names. Indeed, if a proposition is considered to be a statement with a truth value, then any proposition is just the name of either the true or the false. — Banno

Yes, that’s how I understand it too. And for Frege, names are, in an important sense, outside logic, because logic only deals with things with “judgeable content.” Julian Roberts, to whom I often turn for insights into Frege, puts it this way: “Logic [according to Frege] has nothing to say about anything that is not a statement [i.e.], something that can be true or false. The utterance ‛house’, for example [or ‛Berlin’ in my example], is not a statement and cannot be of concern to logic.”

So far as existence is defined, it is defined in terms of the universal quantifier. — Banno

Good. I’m only vaguely aware of how Frege’s concavity differs from our modern universal quantifier, but I think I can still ask these next questions. The Fregean universal quantifier ranges over objects as well as formulae/functions, right? So maybe my question about “Frege on the Beach” (sounds like a hit song) could be phrased as: "When you comprehend the term ‛Berlin’ (as I’m going to assume you do, Herr Frege), does your comprehension depend on the universal-quantification symbol? And then, should you choose to use the term to fulfill a function, does the existence commitment change to ∃x?” So the idea is that ∀ can range over names or terms that are not (yet) part of functions. Clearly, I’m trying to find a way to make a name (and its sense) a “thinkable thought” without violating Frege’s understanding of what logic can do. Does this sound at all sensible to you?

Unless these philosophers explain WHY thought MUST reflect reality (via "logic"), it doesn't seem to have any force to me, except as, ironically, unsupported assertions. — schopenhauer1

One of the best and most accessible parts of T&B is Kimhi's discussion of the four ways we can think about the relation of logic to thought (and reality). I wish I could refer you to this but I don't think it's online anywhere. He tries to develop an argument that, if successful, would answer your request. Oversimplified, here's the idea: If thought does not, necessarily, reflect reality via logic, that gives us a version of psychologism. Thought is just that -- merely thought. And logic represents the way we think, ideally. Kimhi agrees with Frege that this is a non-starter; he calls it "self-refuting":

It cannot ground the normativity of the logical requirements that underlie the objective purport of every claim about how things are -- including the claims of psycho-logicism itself.

In other words, it can't explain the very link you regard as so important -- that between logic and truth about the world. Kimhi goes on to talk about why the thought/world dualism isn't satisfactory either, but I don't want to paraphrase him to death. I hope this gives you a taste of how he justifies his views.

↪Banno

The Parmenidean problem doesn't do much for me either, and I'm sorry Kimhi chose to start his book with it. I'm pretty sure that the majority of the important issues in the book, including my "challenge about assertion", can be addressed without worrying about "saying what is not." And having written that, I'm ready to be schooled by our Aristotelian friends on why this is wrong!

But that's the thing about philosophy, isn't it? I don't see the force of Parmenides' questions. But some excellent philosophers do. So I have three options: 1) I can assume I understand the issues involved, and I am right, and they are wrong. 2) I can reflect harder, assuming that Kimhi is smarter and better read than I am, and he wouldn't be wasting his time on this if there weren't something there. 3) Since the questions don't really interest me, and my time on earth is finite, I can choose to philosophize about something else.

Option 3, thus far, has worked for me.

(Do I really have to go back to the Tractatus? Whimper. OK, I'll take your suggestion.)

@banno @frank

For in a definition certain characteristics would have to be stated. And in application to any particular case the question would always arise whether it were true that the characteristics
were present. So one goes round in a circle. Consequently, it is probable that the content of the word " true " is unique and indefinable — The Thought: A Logical Enquiry

Hmm. Does this amount to pointing out that any definition of “truth” would have to be true, thus opening up the regress? Or is it rather that, armed with our definition of “truth,” we need to use that very definition in order to decide whether the word applies in a particular case? The first option does seem circular to me. But why is second impermissible? Couldn’t I just say, “Yes, it is true that the characteristics that I claim define ‛truth’ are present in this case”? Showing that my definition of “truth” is correct is not the same thing as being able to develop and use it. Or is it? Is there something unique about “truth” that limits “definition” to “correct definition”?

frank

Does this amount to pointing out that any definition of “truth” would have to be true, thus opening up the regress? — J

No. It's basically that truth is integral to the act of assertion. If you give a lecture explaining what truth is, and I'm your audience, I have to already understand what truth is in order to discern what you're doing, that is, telling me the truth about truth. Therefore you can't teach it to me. It's part of my computer's BIOS. It's in my firmware.

I'll see if I can start a thread to go through it.

frank

↪J

Maybe better than starting a thread, would you want to log into the Internet Archive and read some pages from Scott Soames' book on truth?

It's a safe website. Once you log in, you can borrow books. The pages relevant to Frege start on page 21. The previous section covers some issues about truth bearers: sentences versus propositions. That's fairly important for understanding Frege.

Leontiskos

From the other side, once you "grasp" the truth of a situation, have you any choice but to affirm it? This would seem to be somewhat closer to the sense of assertion intended. In other words, Moore's paradox is a simple impossibility: to see the truth of a situation or a proposition is to believe it. — Srap Tasmaner

That seems right to me.

But I want to make a bit of a different point. On my (admittedly limited) understanding of the Tractatus, one thing a picture is entirely incapable of depicting is that it is true. A picture can show how things might be, and things may indeed be that way, but the picture cannot include itself in its depiction and vouch for its own accuracy.

Just so, my belief that a picture is accurate does not count as evidence that it is. — Srap Tasmaner

Okay, and I agree that this seems accurate.

We have talked some -- whether we should have or not, I'm not sure --- about whether there's some sense in which propositions are self-asserting. In these terms, whether a picture at least inherently claims that things stand as it shows, even if it cannot itself substantiate that claim. — Srap Tasmaner

That seems to be what Kimhi is saying:

After a close read of Kimhi's section 2.5 I see him doing the exact same thing that Srap and I were doing. He says that both a declarative sentence and an asserted declarative sentence "display" force, but whereas the first is "a mere display" the second is "a self-identifying display." — Leontiskos

So that we then have:

...what is Kimhi trying to argue here?

Simply working from Boynton, he seems to be saying that if the declarative sentence, "The cloud is raining," does not display an assertion, then it is impossible to say under what conditions the cloud would be raining or not raining. That if "the cloud is raining" and "the cloud is not raining" do not display different assertions, then they could not be associated with different truthmakers. This looks like the same argument that Wittgenstein gives in Rombout 60-62. — Leontiskos

So for Kimhi it is more than a matter of possible worlds. For Rombout this is also true in some sense for Wittgenstein:

That is what Wittgenstein tries to show by means of this analogy: the diﬀerence between (Wittgenstein’s conception of) Fregean assertion and Wittgenstein’s own idea of a sentence saying that its Sinn_w is true. In Tractatus a sentence cannot take a meta-position and compare the Sinn_w it shows with the world. Therefore, the sentence cannot acknowledge its own truth. A sentence cannot ‘check’ whether it corresponds to the world and decide upon the truth of the claim it makes, there are no extra-linguistic acts attributed to sentences. Nevertheless, every sentence aims at truth. A sentence is a picture of reality and as such it already makes a claim about what reality looks like. That is what 4.022 expresses: a sentence depicts a situation and says that this situation is the case. Sentences are not completely neutral phenomena; in showing a situation, they already ‘support’ that situation. It would be incorrect to ascribe intentions to sentences, but a sentence depicts a possibility and embodies that possibility in the same time: the possibility may in fact not obtain, but the ‘proposal’ the sentence makes is that it does obtain. In representing a possibility, the sentence more or less puts the truth of this possibility under consideration.

<<4.2 The meaning of the sentence is its agreement or disagreement of the possibilities of the existence and non-existence of states of aﬀairs.>>

It is not asserted in the Fregean sense, for this agreement is not assertion. Whether it is actually true, or whether it corresponds to reality, seems to be irrelevant to Wittgenstein, at least for logical considerations. It may be part of the natural sciences or of interest to someone, but it is not relevant when formulating the limits of our thought. The actual truth of a sentence can only be established by a subject, and it is already mentioned that a subject does not belong to logic. — Frege, Russell and Wittgenstein on the Judgment Stroke, by Floor Rombout, 63

Leontiskos

As I understand it, 0th level terms - a,b,x - name individuals. Included amongst those individuals are propositions, which in turn name a truth value. So (2+2=4)=(5-3=2)= ⊤. These all name the same individual. So "Berlin" might name the city, and "2+2=5" might name the false.

...

On this account, both "Berlin" and "2+2=4" are names. — Banno

This is close, but it obscures the fact that "Berlin" correlates to "2+2" (or "4") and "Berlin is a city" correlates to "2+2=4." The latter two have truth values, and can be called names in reference to "the True." If I recall, in the source that Bongo Fury gave, on the pages following the citation he gave, Anscombe critiques this understanding of number.

Now does Frege think that "4" and "2+2=4" are both 0th level terms, as Banno implies? I actually don't think he does, but I suppose it's possible. He certainly understands that something which saturates a predicate like "...is a city," differs from something which "saturates" the predicate, "...is the True." He distinguishes between objects and sentences, even if sentences are objects vis-a-vis the True in a certain sense.

Leontiskos

The judgment stoke occurs once in the expression, at the beginning. It affirms the whole expression, not each individual line separately. — Banno

It can only occur once for an expression, but the question is whether a modus ponens is a single expression. What you said earlier simply does not hold:

Frege doesn't write
⊢p⊃q
⊢p
⊢q
such that each is within it's own intensional bracket; he writes
⊢(
p⊃q
p
q) — Banno

My bolding. — Banno

As usual, you are failing to understand the sources you cite. "The judgment-stroke applies to the whole expression, therefore Frege does not use three judgment strokes in a modus ponens." This is a non sequitur.

Here is Rombout, building from my post <here> and Bongo Fury's insight:

Instead of writing the whole inference, consisting of the three assertions “ ⊢ p”, “ ⊢ (p ⊃ q)” and “ ⊢ q” , Russell and Whitehead propose an abbreviation containing the assertions of the two atomic propositions connected by an implication: “ ⊢ p ⊃⊢ q”. Frege would consider this a category mistake; in the Begriﬀsschrift it is not possible to have a judgment stroke within the scope of a conditional.

...

A reason why Russell and Whitehead consider this abbreviation acceptable can be found in their explanation of syllogisms:

<<It should be observed that syllogisms are traditionally expressed with ‘therefore’, as if they asserted both premises and conclusion. This is, of course, merely a slipshod way of speaking, since what is really asserted is only the connection of premises with conclusion.[36]>>

Where in Frege the premises and the the conclusion, as well as the connection between them need to be asserted in order to constitute an inference, this demand is dropped [by Russell and Whitehead]. What is asserted in [Russell's] syllogism is the connection between premises and the conclusion, not the sentences themselves. This seems to be an explanation for allowing for an abbreviated form, but in order to conclude so, it has to be considered whether a syllogism is an inference. — Rombout, 44-5

One can speak about modus ponens in terms of logical consequence or logical inference. Both make sense in their own context, but Frege actually adopts the latter approach. For Frege a modus ponens requires three judgment-strokes.

You continue unabated in your conflation of Frege and your own approach. You are making the exact same interpretational errors that Russell and then Wittgenstein made regarding Frege.

frank

"Berlin" correlates to "2+2" (or "4") and "Berlin is a city" correlates to "2+2=4." — Leontiskos

You need context to make that distinction. If I ask you what the weirdest city in the world is, and you say, "Berlin", you have expressed a proposition. It's all about context.

Leontiskos

'assertoric force follows if a proposition expressed by an indicative clause is presented as relevant in a context made up of a subset of the hearer's factual assumptions." — frank

Who are you quoting? Certainly not Frege. Assertoric force does not depend on the hearer.

frank

Who are you quoting? Certainly not Frege. Assertoric force does not depend on the hearer. — Leontiskos

Please quote Frege on this issue.

Leontiskos

I feel like I'm stuck in an Abbot & Costello routine! If this really represents what Frege would say to me when I ask him whether he comprehends the word printed on that slip of paper he found on the beach, I could only reply, "Well, yeah, but Herr Frege, I'm not asking about extension or objects or what's permissible or impermissible in your philosophy. Have pity on a fellow beachcomber and just tell me whether you understand the word on the paper or not." — J

It is an Abbot & Costello routine. You've been given my testimony, Srap's testimony, Kimhi's testimony, and the testimony of Frege himself, but you still refuse to believe. On what grounds? Frege will not find meaning where there is no extension.

The answer here is that Frege reads the word through his understanding, and his understanding of meaning and truth has to do with the extension of predicates. It makes no sense for you to ask Frege what he would say about a piece of language without appealing to his understanding of language.

Srap Tasmaner

↪Leontiskos

My memory of the Tractatus is that Wittgenstein also says that a picture is a fact, it's part of the world, and so to say this is a picture of that is to say that between this fact and this other fact there is a special picturing relation. And he wants to figure out what it means to say two facts are so related and how it is possible for them to be. (I'm thinking of this in part because Kimhi ends up talking about consciousness, right? So you might take consciousness as the locus of this relation, or even as constituted by it.)

The more natural move for later Wittgenstein is to say that sometimes we see something as a picture of something else, sometimes we don't, and that it is a far more flexible business than it might seem in the Tractatus. The duck-rabbit shows flexibility in what something is a picture of; the possibilities for what counts, or what someone might count, as a picture of a given thing are too numerous to contemplate. (Think of how often only a parent can understand the first indistinct words of their own child. Now consider the range of possible pictures of a dog, from that toddler's wobbly shapes to a cubist painting or an abstract sculpture, and "everything in between", photos, icons, words, pantomime, on and on. Clouds!)

A challenge to Frege on assertion

Welcome to The Philosophy Forum!

Categories

More Discussions