On Explication (Tarski, Spinoza, Leitgeb...)

...[T]he explication of philosophical concepts. That is: the construction of a new concept C′ that replaces a given, often pre-theoretic concept C of philosophical interest, so that the extension of C′ coincides with that of C in the clear-cut and uncontroversial cases but C′ is permitted to, and indeed ought to, improve upon C in terms of exactness, fruitfulness, and simplicity in all the unclear or fuzzy cases. (Leitgeb, 271)**

A true idea must agree with its object--Spinoza Ethics 1A6 (translated by Curley)

"Nor is semantics a device for establishing that everyone except the speaker and his friends are speaking nonsense."--Tarski (1944, 345). [Thanks to Doug Patterson for sending me back to Tarski.]

One of the shared peculiarities of the now canonical philosophers writing between, say, Descartes and Hume is their mistrust of language. They warn of the "danger of being decived by" words, and recommend focus on -- in the memorable words of Berkeley, -- "naked, undisguised ideas." In their terminology, ideas are the objects of the mind. In the hands of Spinoza the distrust means that language -- the play of signs -- is not the vehicle by which to convey truth (E2p41).* This mistrust of language raises many complicated questions about how we  should understand the status of early modern texts (and the rhetorics of writing and reading), and, say, the early moderns' attitude(s) toward revealed writings, but here I focus on a more technical issue: the status of truth.

If language -- read, written, or spoken -- does not partake of the truth, then Tarski's idea that "we must always relate the idea of truth, like the idea of a sentence, to a specific language" (342) must seem like a category error. (The problem does not go away if we focus on functions.) Of course, Tarski does immediately allow that the notion of truth can be extended to other "kinds of objects." Now, some early moderns, not as radical as Spinoza, will allow that truth does pertain  to propositions and/or God's language (or some other ideal language), and so their approaches can be assimilated under Tarski-style explication. 

 

From the semantic point of view of Tarski, Spinoza is so radical that he not just denies the idea that truth is related to language, but also denies that truth pertains to some correspondence between the world and some description or image of it. [Tarski's approach is naturally, but not necessarily, read as a species of the correspondence theory. (343) But nothing turns on this here.] Crucially, in Spinoza there is no gap between between existence and truth. Now, this denial of correspondence is not obvious when a modern reader encounters Axiom 6 (quoted above) at the start of the Ethics. But that axiom essentially says that a true idea must agree with itself. (Recall objects are in the mind.) This is clearer in the Latin: Idea vera debet cum suo ideato convenire. A true idea must coincide with that which it represents. As I have suggested before, adequacy, which when defined (intrinsically) in the Ethics (E2D4), is explicitly contrasted with a correspondence account (Per ideam adæquatam intelligo ideam quæ quatenus in se sine relatione ad objectum consideratur). Adequate and clear ideas are the true.

 

While Spinoza expresses all of this in his philosophical vocabulary, his position, which I will call a 'metaphysical identity theory of truth' is not uncommon among Platonizing philosophers that do not recoil from mysticism (and, perhaps, intellectualized mystics in other traditions). This metaphysical identity theory of truth is also a characteristic of the non-mystical parts of the Hebrew Bible (recall). Given the extended influence of both Platonism and the Hebrew Bible on ordinary life, I would not be surprised if the metaphysical identity theory of truth was reflected in folk practices somewhere, sometimes.

 

Okay, now, let's turn to explication. Explication is great. The partisans of formal philosophy are right to advocate explication. But it is important to be cautious in characterizing what explication is.  In the passage above, which admittedly is a mildly polemical (and celebratory) context, Leitgeb is incautious in a few respects. (I will use his terminology here.) In particular, we should not think of C as pre-theoretical. (Tarski is happy to be taken to explicate the very theoretical "classical Aristotelian conception of truth.") C is not formal (in the modern sense), but prior to Tarski's explication there are lots of (to adopt some terminology from Foucault) 'regimes of truth' that are highly theoretical, even if not formalized with modern logical apparatus. [UPDATE: Actually, as Leitgeb pointed out to me, he was more cautious than I made it seem--I had overlooked HIS "often."] If I am right about Spinoza (and the Hebrew Bible), not all of these 'regimes of truth' share in the "clear-cut and uncontroversial cases."

 

Now, Tarski has a response to the argument of the previous paragraph: "it is undoubtedly the case that in philosophical discussions -- and perhaps also in everyday usage -- some incipient conceptions of this notion can be found that differ essentially from the classical one (of which the semantic conception is but a modernized form)." (355) In fact, Tarski offers what I take to be the prototypical response to this observation: we are dealing "with several different concepts which are denoted by one word." (This is, in part, how he shrugs off empirical survey evidence that the folk conception may deviate from Aristotelian conception.)

 

But there is also a sense that Tarski's response, which (amazingly enough) he dubs "the only rational approach," misses the point. It is simply false that other approaches are not "intelligible and unequivocal." (356)*** Not all regimes of truth are amenable to his strategy of taking a "natural language (or a portion of it in which we are interested) by one whose structure is exactly specified, and which diverges from the given language "as little as possible."" (347) 

 

So, sometimes the purportedly "uncontroversial coincidence" between C’ and C is overstated by the partisans of formality; or, rather, what is taken to be "uncontroversial coincidence" becomes central (and tacitly assumed) to the projects generated C’, but it need not be central to C. This centrality is often projected backwards onto C. In particular, in every explication we run the risk that special features of C, which are invisible in C’, are systematically overlooked or black-boxed in the practice of the technical apparatus developed around C'. To put this in Tarski's terms, C may be home to multiple partially overlapping concepts. (Note that in Spinoza's approach truths entail each other nicely, so it is not like it has only puzzling or unattractive features.) That is to say, the purported clarity of an explication may sometimes conceal what has been lost, especially if it is accompanied without any warnings about the very possibility of loss and with a rhetoric of rationality and the celebration of various scientific virtues. In my view this point generalizes to any formal displacement of experience.

 

I do not mean to be picking on Leitgeb. He is quite frank about possible abuses of formal methods. Perhaps he can inform me of formal methods that do have resources to keep track of the features of a target like C that do not enter into the formalization as C'. 

 

 

*To be clear: while language is always inadequate and confused , it need not be false.

**Today's post is dedicated to Hannes Leitgeb. (I am revisiting an earlier related post on the nature of scientific philosophy.)

***There are also more complicated issues here because Tarski also shows that a rejection of semantic conception of truth runs the risk of embracing contradictions. This is not the place to explore how a Spinozist could respond.

Comments

[Jon Cogburn]

Fantastic post. A few thoughts:

(1) I'm not sure about "Now, some early moderns, not as radical as Spinoza, will allow that truth does pertain to propositions and/or God's language (or some other ideal language), and so their approaches can be assimilated under Tarski-style explication." Maybe in some broad sense, but I think possibly not with Tarski's own framework because the limitations on expressibility might be taken as limits on God's infinitude.

(2) I think Heidegger's critique of the correspondence theory is interesting in this light. Correspondence is fine as far as it goes, but it's not explanatory because the truth of a proposition according to the correspondence theory presupposes (1) that the proposition corresponds to some possible fact, and (2) that that fact be actual.

The actuality of the relevant facts is the deeper notion of truth that has been pushed aside in the post-Tarskian tradition, and ignoring this notion of truth leads to all sorts of problems, including why so many contemporary philosophers to be unable to make sense of how creatures without language (including adult human aphasics and languageless deaf people who are successful in every respect other than language comprehension and production) can even have beliefs.

Heidegger's "aletheia" is unfortunately a step backwards though because it conflates the actuality of facts with the actuality of facts for Dasein. I think it's no accident that "the turn" follows immediately after the truth essay, but he never really managed to really free himself from the correlation.*

Much better to go with Spinoza!

[*Some people working on the Beitraege will disagree with me about this. The jury is out.]

[Eric Schliesser]

Thank you, Cogburn, for your kind words!
On your first (1); yes, there are interesting onto-theistic issues in the vicinity here.
I have wondered -- in my brief engagement with Heidegger -- why Heidegger never explores Spinozistic alternatives to his reading of modernity.

[Nastassja Pugliese]

Great post! Loved it.

I specially like the resulting criticism that if language is not the vehicle of truth, then the semantic theory of truth seem like a categorical mistake. But I tend to think (with, e.g. Zourabichvili) that Spinoza is not as radical in rejecting language as you picture, for it seems that language is the vehicle for truth when it presents the order and connections of ideas and things in nature. His notion of language as a vehicle, of course, is fundamentally different from Tarski's and is much more complicated in S given that truth is its own standard. So, elaborating on this comparison, I tend to think that S's theory of truth is not incompatible with semantic theory. Spinoza is cautionary, and I understand that his goal is to call attention to the epistemic mistake of considering the derivative semantic conception of truth as a substitute of the more fundamental "metaphysical identity theory of truth" (as you named). Both theories, however, can well coexist. Thoughts?

[Eric Schliesser]

Thank you for your kind words.
I agree that Spinoza does not claim that language necessarily leads to falsehood. But Zourabichvili overstates the positive possibilities of language.
Either way, as you claim if language is the vehicle than Tarski can accomodate Spinoza's approach. (And then Tarski's approac has a lot of advantages over Spinoza's.)
However, I do not think that say Spinozistic intuition or the grasping of formal essences is linguistic in any sense.

[Jon Cogburn]

It's a weird counterfactual to think how Spinoza would have reacted to Tarski. I do know the German Idealists would have seen Tarski style solutions to semantic paradoxes just as negatively as Graham Priest does (since Kant's response to the dialectic is the spring from which Tarski on semantic paradoxes as well as the standard set-theoretic response to Russell's Paradox derive). And I don't think the German Idealists would have loved Spinoza if they hadn't seen him (rightly or wrongly!) as anything but antithetical to Kantian/Tarskian finitude.

On Heidegger and Spinoza- Some day I want to write a book called "Martin Heidegger: The Good, The Bad, and the Ugly." The Good is his critique of representationalism and representationalist strains in Southwest school neo-Kantianism. The bad is the neo-Kantian anthropocentrism he got from Husserl and (despite his best efforts) never really managed to shake. The ugly is the blood and soil German Romanticism that occasions both his Nazism and the awful later stuff about history, language, and epochs of being.

[Eric Schliesser]

Well, notice that I had motivated Spinoza as a possible alternative to Tarski not by way of the paradoxes, but by way of their understanding of the nature of truth (and existence).
Having said that, I do think we can show that Spinoza's own approach to truth is motivated by an aversion to certain paradoxes (in particular, concern over Zeno-style paradoxes and a variety of paradoxes pertaining to infinity).
If the bad and the ugly outweigh the good, why write that book?

[Mark Lance]

I didn't' understand what you are calling the "identity theory". Really, just didn't get what the theory is supposed to be. A true idea must agree with itself. Leaving aside that I don't know what it is for an idea - as opposed to a judgment - to be true, why wouldn't a false idea agree with itself just as much.

Anyway, on explication: I'm not sure how serious the objection to Leitgeb is here. The objection seems to be that he is presupposing either that the concept C is pre-theoretical, or that it is uncontested. OK, neither of those should be assumed. (And he doesn't as you note assume the former really.) But these seem to me to require only a mild refinement of what we are doing. We have some prior uses of a bit of language - or if you like of mental items that dispose us to perform speech acts, whatever you want to call them. These may be theoretically infused in various ways, and the uses may or may not be consistent. Maybe there are some core cases, maybe straight ambiguity, maybe hot contest on other cases, or maybe just vagueness. But lets say that there are various structures of norms n1...nm for the use of C. In any event, to explicate for a particular explanatory purpose in the context of a theoretical and methodological tradition, is to construct a new structure of use norms N' for C that bears important family resemblence to central cases of the use of C, and which improves on any of the n's in terms of exactness, fruitfulness, simplicity, etc. relative to the explanatory purposes and context in question.
I think this is pretty much Tarski's idea. I don't think he was trying to make any big claims about all uses of truth. He wanted a use fit for a particular formal context and set of questions that arose in the analysis of formal languages. He precisified existing notions by creating the concept of Tarskian truth. And he thought it maximally fruitful, exact, and explanatory in that context.

[Eric Schliesser]

Let me respond to your second paragraph first. I agree that the objection is not very important. And I agree that you give the right sort of response to it on behalf of Leitgeb. I would say, though, that (a) what seems 'central' is often a post-hoc construct (influenced by the explication); (b) that there ought to be a kind of regulative ideal to try to explicate all features of C, once one starts explicating any one of them. Now, (c) I also believe that it is very possible that there may be features of C that matter a lot in experience yet remain resistant to explication.
On the third paragraph. I think Tarski is more than happy to admit that (i) there are different kinds of uses truth some of which his explication does not capture, and (ii) that alternative approaches may be even more important. Having said that, there is another strain in Tarski that takes a harder line to alternative approaches. (I quoted the line about about his being the only rational approach; other approaches being unintelligible; etc.) [I also think that there are ways of *not endorsing* the semantic conception of truth without embracing contradictions.]
I'll get back to you on the first paragraph because it raises more complicated issues.

[Eric Schliesser]

On your first paragraph. I think you raise two issues: a general concern about what the metaphysical identity theory of truth just is, and an objection. Let me deal with the objection first: it would be tempting to respond that the way false ideas agree with themselves is wrong (because false). But I think this is not quite right. Recall (from teaching, say, the Meditations) that false ideas are confused (unclear, etc.); this means, in part, that they are not properly distinct. More radically, they lack what we would call genuine identity conditions because they are, to use the traditional terminology, privations or (partial) negations. So, the answer to your objection is something like 'false ideas are not really identical to themselves because they are not the kind of things that can be identical to anything.'
On the general concern. Let me think about how to present it. These days ot's obviously very unusual to regularly encounter a view in which truth applies to objects and persons (rather than things said, thought, propositioned about objects/persons, etc.). I think it becomes clear that Spinoza has something like this approach in mind when one grapples with his material on the mind's eternality in Book 5 of the Ethics, which often is not taken seriously by commentators.

[Mark Lance]

Eric: COmpletely agree with (a). I think this is an important point and a virtue. Explication, as I see it ad I think Leitgeb also, is part and parcel of theorizing. We build concepts as we make claims using them - and can't really treat either of these as autonomous - so certainly we are trying to better understand the conceptual terrain as we do it. (b) Why? No one could believe that we ought to explicate all functioning terms. So why does the utility of explicating some uses of C mean we ought to explicate all? Just seems like a case by case matter to me. Rememberr how many really useful words are used in truly idiotic and destructive ways by lots and lots of people. (C) probably.

I think the comment about "the only rational approach" was meant to implicitly include - for the purposes I'm interested in. He thought it really was the only way to explicate truth in formal contexts. He was certainly wrong. Kripke's theory of truth and all the work since just shows that. But this doesn't tell us anything about hte nature of explication or his views on it.

I'm sort of getting my head around this idea theory. I'm really confused by an idea that isn't identical to itself. Why doesn't that mean it doesn't exist. But more generally, would anything be lost if you used a word other than truth here? It sounds to me like a theory of adequate or non-defective ideas. That is, of ideas fit for expressing truths about the world. that's a fine thing to have a theory of and I'll be happy to admit that at some point people used the word 'truth' for it. But that seems like a case where it is really a just plain different notion from the current one. Maybe there is also a view int he offing that this notion of adequacy of ideas can ground the notion of truth of propositions or sentences that others make use of. Anyway, you'll not be surprised that I don't like this sort of theory, but I think I'm getting what the idea is.

[Eric Schliesser]

Well, I admit that there are good reasons to distrust the regulative ideal I propose. But there is a tendency for an explication C' to displace C in the minds of theorists, who tend to forget that C may still have all kinds of role(s) to play in real life and even in theorizing. So, it is important to find ways to keep theorists fully engaged with the messiness in C, one way to do that is to encourage ongoing explications--rather than letting one explication become privileged.

Yes, there is a sense in which a confused idea doesn't really exist.

As I said in the post, it is very tempting just to say that there is a different concept in Spinoza (or, if I am wrong about the historical Spinoza, Spinoza*) that just happens to be called 'truth.' But given the centrality of Spinoza to the history of metaphysics and, I suspect, any thoroughgoing revival of it, I think it would be a mistake to decide this prematurely. Having said that, it is also clear that this alternative approach to truth has some serious drawbacks, especially if you are interested in applications of mathematics.

[Jon Cogburn]

Eric,

Well I probably won't ever write that book, but I should have been clearer that I think that just as far as the contribution to the on-going dialectic (in the strongest possible Hegelian sense), it seems to me that Heidegger's good far outweighs the bad and the ugly. But this is probably just an expression of optimism on my part.

[Aaron Lercher]

Eric linked to this 4/24/18, sending me back in time to here.

Tait's theory of mathematical truth is a contemporary version of the sort of truth concept that Spinoza had, and Descartes also. A proposition is the type of all its rigorous deductions (formulas as types). Tait expands on this by saying how humans obtain warrants by verifying that such deductions or constructions exist. The deductions (which are really mathematical functions) are ideal or abstract objects, while verifications occur in space and time, and are made by computers or humans.

Summing up, truth is the existence of a type of object.

It's not clear how to talk about applied math, as you say. Tait has views on this, though.