Yes, a little more, but first let me be sure I understand you. In 'If p then q', are you saying that 'p' and'q' are mentioned?

The conclusion I suspect is too strong. I'm not keen in including "necessarily". Seems as what is needed is just to be able to set the force to one side in order to consider the propositional content. — Banno

I'll put my original Step 2 here so we can have it in front of us:

Therefore, a proposition cannot contain assertoric force as part of its logical structure. If it could, then we would no longer be able to recognize repetitions of ‛p’ as “the same”: Some would be asserted, some would not, and that would be internal to the structure. There would be some sort of deep or semantic assertion built into this Uber-proposition. Assertion would be functional in the Fregean sense. And since the self-identity of ‛p’ is critical to the entire logical apparatus, we know this cannot be. — J

Does this amount to an argument for the necessity of Frege's separation? I thought it did, on the grounds stated: We've got to have all repetitions of 'p' as identical, but if some of them have assertions "built in," then it wreaks havoc with synonymy, inference, etc., as Step 1 seems to show. But perhaps my conception here is too narrow.

I don't get why the "judgement stroke" shown at the beginning of the claim matters so much to Kimhi.. What am I not getting.. This is almost the definition of minutia-mongering to me..

"The grass is green", and rephrasing it in another way whereby it is indicated by a symbol that this is a judgement, seems like a bafflingly tenuous thing to take umbrage to.

If his point is that propositional logic takes the context out of the equation and thus makes the subtleties of human life distorted due to the logic, okay... But it's simply a tool, and can probably be used in a number of manners that can make it fit. But maybe I am not seeing how insidiously wrong Frege's analysis tool is..

@banno @leontiskos @srap tasmaner @janus @fdrake @schopenhauer1 @russellA. and apologies to anyone I missed:

@Pierre-Normand has found a very good short review of Thinking and Being here: unpublished review essay by Owen Boynton. I actually have a lot to say about it but time is tight, which brings me to the other thing I wanted to say:

I don't know if others have this problem, but I always feel like I ought to reply to interlocutors on an interesting thread as quickly as possible, maybe especially when it was my OP. Part of me knows there's no rule about this, and the fact is that I often can't because I have this Other Life where I have to go places and do things. No big deal, of course -- none of us is so terribly important to the others -- but I thought I'd just say it out loud. Right now, there are a number of points I'm eager to address but you won't be hearing from me till tomorrow at the earliest.

Agreed. Good analysis. I'd only add that whether there is indeed a "wholeness of KG" is a central question, and Kimhi is trying very hard to argue for it, using pre-Fregean concepts of logic. — J

Yes, that's right. But it seems clear enough to me that the assertoric nature of KG is different from the assertoric nature of FG. That is where I think one should begin, and it moves us into the syncategorematic question.

---

minutia-mongering — schopenhauer1

That's an understandable read. But as Aristotle noted, small errors do add up over time. Kimhi is thinking in terms of decades and centuries.

But [logic is] simply a tool — schopenhauer1

Not for Frege or Kimhi (or Aristotle). If Kimhi or Frege thought logic were just a tool or an approximation or a pragmatic matter, then Kimhi's book would be completely moot.

To be fair, Aristotle would probably admit that his "syllogistic" maps human inference only imperfectly, but if you read that syllogistic in context it is not meant to be self-supporting.

---

Frege, and logic, moved from prefixing"I know..." to something more like "I can write..." over time. — Banno

You are still conflating Frege with your own approach and a post-Fregian trend in logic. Frege never does this. See Rombout, sections 2.1.2 and 2.1.3. Frege never prefers consequence to inference. In fact he explicitly opposes such a move. For Frege such an approach is, "playing with mere words."

Satisfaction, and so to a great extent truth, enter into the process if at all at the level of interpretation. — Banno

Not for Frege. Not for Kimhi. See, for example, the abstract of, "Truth and Satisfaction: Frege Versus Tarski."

Not for Frege or Kimhi (or Aristotle). If Kimhi or Frege thought logic were just a tool or an approximation or a pragmatic matter, then Kimhi's book would be completely moot.

To be fair, Aristotle would probably admit that his "syllogistic" maps human inference only imperfectly, but if you read that syllogistic in context it is not meant to be self-supporting. — Leontiskos

The problem with all this is it seems very much arguing abstractly.. like we are arguing over something that doesn't seem to have a real "center". WHAT exactly is Kimhi proposing.. Logic is part of thinking and being.. Okay, but WHAT does that really MEAN. I get he criticizes Frege for making logic more of a tool of analysis rather than tying it to human "use-contexts", or whathaveyou (or I think that's his complaint), but what of it? What is the real insight here that is profound, insightful, or meaningful? Why would his work matter?

And if it goes back to some more minutia about Frege, I think you have not quite got what it is that would answer this question.

but I thought I'd just say it out loud. Right now, there are a number of points I'm eager to address but you won't be hearing from me till tomorrow at the earliest. — J

That's very kind of you to highlight. I'm not expecting responses personally, and I encourage you - as a mod - to take whatever time you need to respond. Your posts and the discussion you have fostered here are a model in both content and conduct.

Just to be sure, I would agree that "assertoric force may be dissociated from predication" but am disincline to say "assertoric force is necessarily dissociated from predication".

An example. Consider two predications: "Grass is green" and "J believes that Grass is Green". We might perhaps happily remove the assertoric force from the first "grass is green", but not so much from the second. That's because it is within the scope of the belief. It's extensionally opaque.

Now "It is raining" presumably is the same in both.

I don't wish to rule out a logical analysis of such belief statements. So I don't wish to say that the force, in this case a belief, must necessarily be dropped. There's no need for such a broad preemption.

This is by way of leaving open the possibility of say formal treatments of belief, doxastic modal logics, or analysis of belief revision.

And yes, this "wreaks havoc with synonymy", so that Davidson for example splits such belief predicates into "p" and "J believes that p" to bring out this very issue. The repetitions of "p" here are not identical - one is a proposition, the other is a name in a different proposition.

If you like, I would maintain the possibility of treating such utterances logically. So your rule might apply within a first-order logic, and so for Frege, but not for higher order logics in which we predicate with other propositions.

By "mention" I meant simply that you might reference the sentence as we are doing here, without asserting anything about its truth. So "mention" refers to the whole sentence under consideration..

Cheers, and success to your endeavours in the "Other Life". Thanks also for the reference. I have by now read a few such reviews, but all have in common the lack of a clear argument or thesis from Kimhi. It bothers me that no one seems able to set out in a few hundred words what is being argued – I think you might agree with this. There's also a question of what might be called "style", in that so much of the writing on this topic is circuitous and off-point, reminiscent of Hegel rather than of Russell. Despite that I have some sympathy for views apparently expressed in the book. Oddly, the hardcover and kindle editions are nearly the same price. So I still baulk at forking out the money.

Anyway, given my tight wallet and the absence of a clear account of what Irad Kimhi has to say, I've probably finished with my comments. I'll keep an eye on developments.

Who dug this rabbit hole? Lewis Carroll, apparently.

In Principles of Mathematics Russell falls into confusion through a desire to say both that, e.g., 'Peter is a Jew' is the same proposition when it occurs in 'If Peter is a Jew, then Andrew is a Jew', and that it is not. It must be the same, because otherwise modus ponens would not be valid; it cannot be the same, because then 'Peter is a Jew; if Peter is a Jew, Andrew is a Jew; therefore Andrew is a Jew' would be the same as 'If both Peter is a Jew and if Peter is a Jew, then Andrew is a Jew, then Andrew is a Jew', and it was precisely Lewis Carroll's discovery (in 'What the Tortoise said to Achilles') that it was not. Frege provides a solution by saying that the sense of the two occurrences of 'Peter is a Jew' (the thought expressed by them) is the same, but that the assertoric force is present in one and lacking in the other. — Michael Dummett: Frege, Philosophy of Language, page 304

(My emphasis.)

I'm not falling down it. Maybe I need a push?

Yes, 'Peter is a Jew; if Peter is a Jew, Andrew is a Jew; therefore Andrew is a Jew' says that Peter is a Jew.

Whereas, 'If both Peter is a Jew and if Peter is a Jew, then Andrew is a Jew, then Andrew is a Jew' doesn't.

So what? Why deny, in the latter case, that the sub-string 'Peter is a Jew' (considered as such, apart from its context) still says so? You could perfectly well admit that it does but still say the whole, larger sentence doesn't.

And if you have a reason, why shouldn't it equally well apply for sense, and disqualify the inner occurrence of the sentence from having the same sense as a free-standing occurrence?

So your rule might apply within a first-order logic, and so for Frege, but not for higher order logics in which we predicate with other propositions. — Banno

That's all I meant, yes. And of course there are several points of view in this thread, including Kimhi's, that call into question this way of seeing the matter.

OK, but I'm still trying to break it down a bit more. 'If p then q' is a sentence, but so are 'p' and 'q', presumably. To answer your question about "whether there's more to it" (the Frege / Kimhi issue), I need to start by getting clear on how you're using "mention" as opposed, I presume, to "use." Pardon the persnicketiness, but it makes a difference in what I want to go on to say. So . . . which sentence are you referring to as mentioned? (or all three?)

Well, I hope you jump in if the mood strikes you.

It bothers me that no one seems able to set out in a few hundred words what is being argued [in Thinking and Being] – I think you might agree with this. — Banno

I sure do, and also about his style. Where were the editors at Harvard UP?! But let's face it, he has this in common with other great philosophers. Kant's style is appalling, but when you work at it, much is revealed. And if someone asked me to explain the 1st Critique in a few hundred words, I'm sure I couldn't. Kimhi's book is much shorter but I believe the depth and importance is there, so paraphrasing isn't easy. I think the Boynton review does the best job, and I'm going to post a few thoughts about it later.

I am so lost in the sample sentences. Is this really how Dummett presents them? Would you mind either punctuating them differently (parentheses, maybe, to set off the sentences-within-sentences?), or else adding a translation into symbols?

I see your point. (But yes the quote is direct.) Ok,

it cannot be the same, because then

"P; if P then A; therefore A"

would be the same as

"If (P and (if P then A)) then A",

and it was precisely Lewis Carroll's discovery (in "What the Tortoise said to Achilles") that it was not.

I'm not falling down it. Maybe I need a push?

Yes, "P; if P then A; therefore A" says that P. (Asserts "P".)

Whereas, "If (P and (if P then A)) then A" doesn't.

(As long as "P" is understood as shorthand for "Peter is..." etc.)

So what? Why deny, in the latter case, that each occurrence of the sub-string "P" (considered as such, apart from its context) still says that P? You could perfectly well admit that it does but still say the whole, larger sentence doesn't.

And if you have a reason, why shouldn't it equally well apply for sense, and disqualify the inner occurrence of the sentence from having the same sense as a free-standing occurrence?

Yes, you’ve got it, as your later post with the extensive Kimhi quotes shows. Kimhi agrees with what he calls “Frege’s observation” but not what he calls “Frege’s point.” His line of dialogue should read, “I disagree, if you’re saying that the only thing which gives the predicate its force is assertion. But as I read you, you needn’t be saying that at all. That’s a conclusion that Geach and other Fregeans have imposed on you.” And that’s what I’ve been saying too.
— J

So Kimhi doesn't disagree with Frege after all? He only disagrees with Geach? — Leontiskos

No, you’re right, I was oversimplifying for the sake of brevity. He thinks that both Frege and Geach maintain “Frege’s point.” So a better rewrite of the words addressed to Frege that I put in Kimhi’s mouth: “I disagree, if you’re saying that the only thing which gives the predicate its force is assertion. But as I read you, you needn’t be saying that at all. But you do, and Geach and other Fregeans have emphasized this additional point without seeming to realize that you could have stopped with your ‛observation’ and all would have been well.”

I don't think you're grasping the seriousness with which Frege excludes existence as a predicate. My second quote here literally has Frege explaining why it makes no sense to speak about the existence of entities or the non-existence of entities. I don't see how this claim of yours can be saved:

2) we have to start with a logically grammatical proposition that fills the argument slot with a term, thus creating what Frege called a “name,” before we can say whether it exists or not.
— J

If "before" means "before" and "say whether" means "say whether," then Frege will deny this claim. — Leontiskos

I interpret Frege differently here. To show how, let me start with a question. What do you think the status is of the term ‛The grass in my backyard’? Are you able to understand it? And now a second question: What do you think Frege would say?

The concept of the syncategorematic may need to be introduced, even if the word is not. — Leontiskos

I’m starting to think so too, and see Boynton. He does a far better job than I thought possible at giving the term some intuitive appeal, especially when he likens it to “metaphysical”.

That is very kind of you to say, thanks.

The thread has flaws. I freely admit it. But I still think it is a good thread, precisely because it does not easily fit our preconceived categories and is not reducible to the standard tropes of TPF. It is an exploratory thread which favors a kind of thinking together. And at the very least it will generate many interesting thoughts, and probably also interesting threads.

yes the quote is direct — bongo fury

Yikes! Thanks for the translation.

Why deny, in the latter case, that each occurrence of the sub-string "P" (considered as such, apart from its context) still says that P? You could perfectly well admit that it does but still say the whole, larger sentence doesn't. — bongo fury

Well, here we are back to the vexing question of "assertion" a la Kimhi. To push you down the rabbit hole, I'd need to persuade you that your use of "say" and "says" is not innocent, but brings with it an entire apparatus involving what it is for a consciousness to think (and possibly assert) a proposition. I'm still working on the best way to talk about this (and I'm not sure it's true, but Kimhi makes it plausible at least).

And if you have a reason, why shouldn't it equally well apply for sense, and disqualify the inner occurrence of the sentence from having the same sense as a free-standing occurrence? — bongo fury

This is a little easier. We can go along with Frege in the separation of sense from assertion as merely meaning that a concrete occurrence of a proposition may actually assert it, whereas this assertive force is not associated with the repeatable symbol 'p'. ("Frege's observation," above). What is necessarily associated is the sense.

Good post, .

I have a copy of Kimhi’s book on loan and I tried to put it down given the constraints that J laid out in the OP. Of course I picked it up again when the thread began stalling. On picking it up again, I am struck by the fact that it is primarily a Wittgenstenian book. This is true not only because Wittgenstein is the main source, but also because there is a lack of overly clear answers and theses. It is allusive and elusive.

Perhaps Wittgenstenians can comment on Kimhi’s use of Wittgenstein. We could look at a few of the things that Kimhi appeals to in Wittgenstein. I already gave one here:

[11] This dissociation is the target of TLP 4.063, which purports to show that “the verb of a proposition is not ‘is true’ or ‘is false’ as Frege thought: rather that which ‘is true’ must already contain the verb” (Ludwig Wittgenstein, Tractatus Logico-Philosophicus, trans. Pears and McGuiness [Oxford: Blackwell, 2001], 29). — Kimhi, Thinking and Being, 8

(Rombout gives the fuller context and an analysis of this quote on page 60 and following)

A second is footnote 27 from the excerpt given in <this post>:

[27] For Geach, Frege’s observation is meant to be restricted to occurrences of propositions in logical contexts that are identifiable as extensional or truth-functional. But understood as Wittgenstein’s point, Frege’s observation is not limited to a truth-functional context, and it implies semantic innocence, p is the same in p and in A thinks p. An expression of a generalized version of Frege’s observation, one that assimilates intensional and extensional contexts, can be found in the following note from Wittgenstein’s Notes on Logic:

<<When we say A judges that, etc., then we have to mention a whole proposition which A judges. It will not do to mention only its constituents, or its constituents and form but not in the proper order. This shows that a proposition itself must occur in the statement to the effect that it is judged. For instance, however “not- p” may be explained, the question “What is negated?” must have a meaning. (96)>> — Kimhi, Thinking and Being, 39

A third:

Since the subordinate propositions in a compound are treated as logical building blocks, so to speak, I will call this a spatio-logical account of truth-functional propositional complexity.[35]

[35] G. E. M. Anscombe uses the term “logical chemistry” to describe such an account:

<<Consider the explanations of propositions and truth-functions, or logical constants, which are commonly found in logic books. It is usual for us to be told: first, propositions are whatever can be either true or false; second, propositions can be combined in certain ways to form further propositions; and third, in examining these combinations, i.e., in developing the truth-functional calculus, we are not interested in the internal structure of the combined propositions. . . . Is there not an impression as it were of logical chemistry about these explanations? (An Introduction to Wittgenstein’s Tractatus [New York: Harper & Row, 1965], 53)>>

Anscombe correctly rejects a reading of the Tractatus that ascribes to it this “logical chemistry.” My suggestion is that the main features of such an account arise from Frege’s point. — Kimhi, Thinking and Being, 48

A fourth:

The notion that logic is not concerned with actual, historical occurrences of linguistic expressions but only with symbolical occurrences of expressions within larger symbolical contexts lies at the heart of Wittgenstein’s early work. Later he would note, for example, that the common or regular agreement between speakers in what they describe by the use of the predicate “. . . F,” is not logically external to the assertoric act of describing something as “. . . F.” And he insists that, in saying this, we do not lose the integrity of logic—we do not give in to psycho-logicism.[36]

[36] Compare Philosophical Investigations §242:

<<If language is to be a means of communication there must be agreement not only in definition but also (queer as this sounds) in judgments. This seems to abolish logic, but does not do so. It is one thing to describe a method of measurement, and another to obtain and state results of measurement. But what we call “measuring” is partly determined by a certain constancy in results of measurement.>>

The remark seems addressed to his own earlier separation of logic and psychology. — Kimhi, Thinking and Being, 51

@J

So I finished reading Rombout’s thesis, “Frege, Russell and Wittgenstein on the Judgment Stroke” (but I did skip the section on Kant). English is certainly not Rombout's first language, so there are errors of grammar and spelling, but it was on the whole very good. It sheds a lot of light on this issue, and interacts with Kimhi in a very odd way.

The thesis is something like this: “Wittgenstein’s understanding of Frege and especially Frege’s judgment-stroke is flawed. Anscombe thinks this is because Wittgenstein was reading Frege through Russell. Was she right, or was it flawed for some other reason?”

This raises two possibilities with regard to your thread. First, that Kimhi simply followed Wittgenstein in his misunderstanding of Frege. In that case Rombout’s survey of Wittgenstein’s response to Frege inevitably captures Kimhi himself. Either way, that section is very helpful, both in itself; because Kimhi seems to be building on Wittgenstein’s critique; and because the Wittgenstein/Kimhi critique may be successful even if they are both failing to really understand Frege’s position. It also has some overlaps with Banno’s misreading of Frege, probably because Banno is a Wittgenstenian.

The second possibility is that Kimhi rightly understands Frege and avoids Wittgenstein’s misunderstandings, and nevertheless utilizes, augments, and improves Wittgenstein’s critique.

I don’t know which is true: probably some of both. Kimhi includes no mention of the historical debates about Wittgenstein’s interpretation of Frege that are presented by Rombout, and he certainly would have had he known about them. On the other hand, he seems to have a better grasp of Frege than Wittgenstein did. If the first possibility holds then the ‘Frege’ referred to in the OP is really Wittgenstein’s interpretation of Frege, which is a possibility that must be held in mind. My own sense is that Kimhi misunderstands Frege along the same lines as Wittgenstein, but with some mitigation, and in a way that might not undermine his sub-thesis.

I find it a bit odd the way Kimhi places Aristotle and Wittgenstein in the same bed. If Aristotle and (early) Wittgenstein both disagree with Frege, it is for wholly different reasons. And while I find Wittgenstein’s alternative to Frege quite terrible, his critique is nevertheless acute and worthwhile. That critique via Rombout may be a good way to get the thread on track without appealing too strongly to Kimhi. It is found in Rombout 4.3.1, and to a lesser extent in the following sections.

(I should also say that in general it proves difficult to navigate the various logical paradigms, even before Kimhi is brought in. It is also worth checking out PI #22, which Rombout references and Kimhi does not include.)

---

I pointed out where Rombout gives context for Kimhi’s short quote from Wittgenstein, but the same holds for Kimhi’s short quote of Russell:

In grammar, the distinction is that between a verb and a verbal noun, between, say, “A is greater than B” and “A’s being greater than B”. In the first of these the proposition is actually asserted, whereas in the second it is merely considered. But these are psychological terms, whereas the difference which I desire to express is genuinely logical. It is plain that, if I may be allowed to use the word assertion in a non-psychological sense, the proposition “p implies q” asserts an implication, though it does not assert p or q. The p and the q which enter into this proposition are not strictly the same as the p or the q which are separate propositions, at least, if they are true. The question is: How does a proposition differ by being actually true from what it would be as an entity if it were not true? It is plain that true and false propositions alike are entities of a kind, but that true propositions have a quality not belonging to false ones, a quality which, in a non-psychological sense, may be called asserted. There are grave difficulties giving a consistent theory on this, for if assertion would in any way change a proposition, no proposition which can possibly in any context be unasserted could be true, since when asserted it would become a different proposition. But this is plainly false; for in “p implies q” p and q are not asserted, and yet they may be true. Leaving this puzzle to logic, however, we must insist that there is a difference of some kind between asserted and unasserted propositions.

(Russell, Principles, §38) — Rombout, 33

Rombout basically looks at Frege's judgment-stroke, and then at the ways that Russell and Wittgenstein wrestle with Frege's judgment-stroke. It is both helpful and freely available.

Who dug this rabbit hole? Lewis Carroll, apparently.

...

I'm not falling down it. Maybe I need a push?

Yes, 'Peter is a Jew; if Peter is a Jew, Andrew is a Jew; therefore Andrew is a Jew' says that Peter is a Jew.

Whereas, 'If both Peter is a Jew and if Peter is a Jew, then Andrew is a Jew, then Andrew is a Jew' doesn't.

So what? Why deny, in the latter case, that the sub-string 'Peter is a Jew' (considered as such, apart from its context) still says so? You could perfectly well admit that it does but still say the whole, larger sentence doesn't.

And if you have a reason, why shouldn't it equally well apply for sense, and disqualify the inner occurrence of the sentence from having the same sense as a free-standing occurrence? — bongo fury

This is excellent, and it has everything to do with the OP. :up:

See Rombout:

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 a 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

That entire section in Rombout's paper discusses this issue, which seems central to the OP.

a kind of thinking together — Leontiskos

It's ironic you say this.

My deep dissatisfaction with everything I've read of Kimhi was precisely the emphasis on assertion, judgment (a word I've never had any use for because of its libertarian aura), and this "I" of logic.

I've been thinking a lot the last few days about the "we" of logic, but so far it's not in good enough shape for the thread I promised.

Anyway, this "I" stuff is why I'm not bothering about Kimhi anymore.

One way to phrase the problem is as follows:

The issue of two phases in the assertion [of] a sentence is also discussed in Wittgenstein’s Philosophische Untersuchungen [§22]. — Rombout, 61

It's tricky to switch paradigms, but in Wittgenstein's paradigm the problem is that Frege has "two phases in the assertion of a sentence." Russell struggles with the same issue from a different paradigm. For Frege it is the difference between "the True" and the judgment-stroke.

To try to put it plainly: is it possible to see that something is true before going on to assert it? And does (the recognition of?) a sentence's truth require a subject? Is the syncategorematicity (in Boynton's sense) of the judgment-stroke already present in the truth-assessment?

The puzzle is explicit in Frege's requirement that only true sentences can be asserted, a requirement that is incomprehensible to, and thus not even understood by, Russell and Wittgenstein. If only true sentences can be asserted, then what exactly is the difference between calling a sentence true and asserting it? Frege has an uncommonly objective notion of truth (and also assertion) (at least as far as contemporary logic is concerned).

So a better rewrite of the words addressed to Frege that I put in Kimhi’s mouth: “I disagree, if you’re saying that the only thing which gives the predicate its force is assertion. But as I read you, you needn’t be saying that at all. But you do, and Geach and other Fregeans have emphasized this additional point without seeming to realize that you could have stopped with your ‛observation’ and all would have been well.” — J

This seems cumbersome, but certainly better. And now we are correctly speaking about assertoric force rather than non-assertoric force, which was my point. Remember though that Kimhi does not think one can stop with Frege's observation. He thinks we need to go on to draw Wittgenstein's point instead of Frege's point.

I interpret Frege differently here. To show how, let me start with a question. What do you think the status is of the term ‛The grass in my backyard’? Are you able to understand it? And now a second question: What do you think Frege would say? — J

My very first sentences in the thread:

The closest Frege's system can get to modeling something like this is to say, "There exists something which is both grass and green." Fregian logic has an especially hard time with individuals since it is built for concepts or classes. Given that the statement is not Fregian in the first place, it raises a whole host of issues. — Leontiskos

Frege will say, "There is something which is grass, and is in my backyard, and the class of things which are thus and so is not empty." He will not say, "...therefore, backyard-grass exists." He will say, "...therefore the class of backyard-grass is not empty." What is being rejected is the subject-predicate approach, such as Aristotle's where grass is a substance and "in my backyard" is an accident. For Frege it is a matter of applying two functions to a single 'argument', and the argument does not sub-sist as something independent of the functions.

I take it that Novák is required reading at this point, and is deeply related to one of Kimhi's central theses, regarding speech about that which is not.

I’m starting to think so too, and see Boynton. He does a far better job than I thought possible at giving the term some intuitive appeal, especially when he likens it to “metaphysical”. — J

Yes, I agree. I have been meaning to look at Kimhi more closely on this subject.

and the class of things which are thus and so is not empty — Leontiskos

Eek. Not only is this part redundant but it requires classes to be objects, which pill, though bitter, even Quine swallowed for the sake of mathematics. But we don't have to go there just for this.

(It's also Quine who pointed out that names for individuals are eliminable. You just make a predicate like "Socratizes" that is satisfied by a single individual. That might not strike you as either intuitive or felicitous, but it's a typical math move, to subsume a particular problem into a more general one.)

The puzzle is explicit in Frege's requirement that only true sentences can be asserted, a requirement that is incomprehensible to, and thus not even understood by Russell and Wittgenstein. If only true sentences can be asserted, then what exactly is the difference between calling a sentence true and asserting it? Frege has an uncommonly objective notion of truth (and also assertion) (at least as far as contemporary logic is concerned). — Leontiskos

Indeed it is so incomprehensible that I didn't even remember this was Frege's view.

Which suggests to me that "assertion" is really not a word we should be using at all here, given its modern meaning.

I may have missed it, but I suppose this applies to "judgment" as well, that you cannot judge as true what is false.

All of which points toward that favorite (never defined) word, "grasping". So it's about grasping the meaning of a proposition, grasping its truth, the difference between those, and so on.

It's ironic you say this.

My deep dissatisfaction with everything I've read of Kimhi was precisely the emphasis on assertion, judgment (a word I've never had any use for because of its libertarian aura), and this "I" of logic.

I've been thinking a lot the last few days about the "we" of logic, but so far it's not in good enough shape for the thread I promised.

Anyway, this "I" stuff is why I'm not bothering about Kimhi anymore. — Srap Tasmaner

Rombout's section on Kant (2.2.2) seems very related to this issue. Instead of trying to comment further I will await your thread (although I will say that I think something more radical than a shift to "we" is required to displace Frege's presuppositions).

But the thread fosters "thinking together" insofar as the premise is something like, "Frege says this, and Kimhi critiques it. How might we understand Kimhi's critique? What exactly is it that is wrong with Frege's claim?" That creates a common project and thinking together.

But I am now seeing the relevance of some of your earlier comments, such as those about Wittgenstein's picture theory and his emphasis on "showing."

-

What that gets us, I'm not sure. If you say, for instance, that assertion "aims at truth" (which, perhaps mistakenly, I suppose is the sort of thing Kimhi will want to say), then a declarative sentence must be the sort of thing that can be aimed at truth, whatever that means. — Srap Tasmaner

Compare:

“in order to express a thought, I have to realize that thoughts aim at truth" — Michael Potter as quoted in Rombout, 61

I always read the "language-game" analysis as an expansion of the context principle, so I have some sympathy with this view.

I do want to note the alternative approach, though, which is Grice's, and which I also have considerable sympathy with. Grice distinguishes sentence meaning from speaker's meaning, and defends the usual logical analysis of the meaning of a sentence as essentially correct, even if in a given context a speaker means something else by saying it.

An example I've used before: you're driving somewhere with a friend and ask, "Should we stop here to eat?" Your friend checks his phone and says, "The next town is like 70 miles." What he means by saying this is "yes", but that doesn't change the meaning of the sentence he uttered or of any of the words in it. --- Nor is "yes" logically implied by what he said; it is only implicated, and he might in fact be willing to wait.

I say all this because if you want to identify the meaning of a sentence with its use, as a move in a language-game -- what I think Kimhi might be pointing at with "actual occurrences" and so on -- you can get speaker's meaning right but skip entirely over sentence meaning, which in this case is a verifiable claim about geography.

It does seem like the principal subtext here is the picture theory of the Tractatus and its abandonment. — Srap Tasmaner

This seems right to me.

-

Eek. Not only is this part redundant but it requires classes to be objects, which pill, though bitter, even Quine swallowed for the sake of mathematics. But we don't have to go there just for this. — Srap Tasmaner

I agree it is redundant. That has sort of been my point. And maybe Frege would reject it on the basis that classes are not objects, but from what I understand he vacillates on this a bit. I am not averse to the conclusion that he rejects it. That seems most consistent.

(It's also Quine who pointed out that names for individuals are eliminable. You just make a predicate like "Socratizes" that is satisfied by a single individual. That might not strike you as either intuitive or felicitous, but it's a typical math move, to subsume a particular problem into a more general one.) — Srap Tasmaner

Yep:

However, in his reception of Russell’s ideas Quine makes one inconspicuous but crucial modification: any mention of “genuine” proper names is left out, to the effect that all proper names are in fact disguised descriptions. In this way, our language is finally devoid of any means whatsoever to genuinely and uniquely refer to a fixed individual; in Quine’s conception, individuals can only be reached via (quantified) variables. Thus, according to Quine, we are not only denied the capacity to speak of that which is not, but we also cannot, directly and by name, speak of individuals that are: this is manifested e.g. in Quine’s rejecting not just “possible entities”, but de re modalities in general.[13] — Lukáš Novák, Can We Speak About That Which Is Not? Actualism and Possibilism in Analytic Philosophy and Scholasticism, 166

-

Indeed it is so incomprehensible that I didn't even remember this was Frege's view.

Which suggests to me that "assertion" is really not a word we should be using at all here, given its modern meaning. — Srap Tasmaner

I don't find the usage beyond the pale, but I now see how it can be confusing. For example, you are certainly thinking of assertion differently than Frege in passages like this:

One adjustment to this I would probably make is to say the goal of assertion is to aim someone else at truth -- at what you take for truth, anyway, so that's another adjustment. — Srap Tasmaner

For Frege one asserts even when they are working out syllogisms in their room alone. On page 38 and following Rombout has a very interesting discussion of performative aspects of written language, such as the judgment-stroke.

I may have missed it, but I suppose this applies to "judgment" as well, that you cannot judge as true what is false. — Srap Tasmaner

Yes, all we are ultimately talking about is the judgment-stroke.

All of which points toward that favorite (never defined) word, "grasping". So it's about grasping the meaning of a proposition, grasping its truth, the difference between those, and so on. — Srap Tasmaner

Yep!

Broadly -- I think everyone knows this, but here we are -- the two principal strands of thought about language are: language as symbol system (which facilitates thought); language as communications system. Frege is generally treated as part of the former camp, and early Wittgenstein, and the latter camp includes later Wittgenstein, Grice, et al. (David Lewis makes an heroic attempt to marry them in Convention, and admits that he cannot.) For what it's worth, I'm in the latter camp, but see the sort of analysis the former produces as a useful strategy in some cases.

But language is easy (!) compared to logic. It appears to me that research overwhelmingly supports the communication-first view, but there is no simple path from there to a similarly robust take on logic, not that I've found anyway. That's uncomfortable for me, but oh well.

And Kimhi seems to me mostly to be talking about a pretend world, or at least mistaking the simplifications (that is, fictions) we employ, like "grasping the truth of a proposition", for reality. (And note that mathematics is such a pretend world where logic runs like a champ without any of this psychological baggage.)

Broadly -- I think everyone knows this, but here we are -- the two principal strands of thought about language are: language as symbol system (which facilitates thought); language as communications system. Frege is generally treated as part of the former camp, and early Wittgenstein, and the latter camp includes later Wittgenstein, Grice, et al. (David Lewis makes an heroic attempt to marry them in Convention, and admits that he cannot.) For what it's worth, I'm in the latter camp, but see the sort of analysis the former produces as a useful strategy in some cases. — Srap Tasmaner

I'm not convinced that we have to choose, and Rombout's discussion of performative language gets at this starting on page 38. In fact I tend to wonder if Wittgenstein's irrecoverable mistake was choosing and shoehorning language in this way. For Aristotelians natural language is something that can never be fully accounted for or categorized, and I tend to think that this is essential to philosophy. As soon as you try to enforce rules on language you're screwed, even if your rules are true 98% of the time.*

language as symbol system — Srap Tasmaner

It is said that Peirce and Frege created the same system independently, but I am curious whether Peirce's robust focus on semiotics would have produced a better logic than Frege's. Peirce also seems less invested in mathematics, which investment is another kind of rule or determination over language. Aristotle explicitly opposed the mathematization of knowledge and logic.

But language is easy (!) compared to logic. It appears to me that research overwhelmingly supports the communication-first view, but there is no simple path from there to a similarly robust take on logic, not that I've found anyway. That's uncomfortable for me, but oh well. — Srap Tasmaner

I would start with the idea that a symbol or sign is a kind of self-communication. Apparently Russell talked about the usefulness of written symbols along these lines.

And Kimhi seems to me mostly to be talking about a pretend world, or at least mistaking the simplifications (that is, fictions) we employ, like "grasping the truth of a proposition", for reality. — Srap Tasmaner

This is where we disagree, but the disagreement runs so deep that it deserves another thread. :wink:

I may as well respond to this, even if it is better fitted to your new thread:

Shrug. That's how simplification works. It's a model; all models are wrong. — Srap Tasmaner

Frege is not giving a simplification or a model, and it is essential to understand this. Whether it makes sense to say that every conception of logic is necessarily a simplification (a simplification of what?) is something that I think would fit a new thread (or the old Sider thread, which was closely related to these questions).


* See, for example:

One of the interesting points that Novák makes is that there is a characteristic divide between the scholastics and the analytics with respect to natural language: — Leontiskos

In scholasticism the matters are rather more complicated. Generally speaking, the scholastics lacked the Russellian revisionist attitude towards natural language, and therefore they rarely explicitly challenged the obvious capacity of the natural language to refer to non-existents. Their approach was, generally, to explain and analyse, not to correct language – and so the standard scholastic theory of supposition (the mediæval counterpart of reference) naturally allows (via devices like ampliation etc.) for reference to non-existents.[18] — Lukáš Novák, Can We Speak About That Which Is Not?, 168-9