@Isaac Any tertiary sources?

@Banno

Even those thinkers who are certain there is only one conceptual scheme are in the sway of the scheme concept; even monotheists have religion.

But this (as the implied alternative to polytheism) leaves out ecumenism/pluralism, which I think characterises most of the philosophical "persuasions" to which Davidson might be referring, e.g. those countenancing,

• conceptual schemes (by that name, but equally...)
• paradigms
• social constructs
• mental models
• forms of life
• language in use
• semiotics
• world-making

I would guess that it's only a minority of sects that have believed in a clear separation of one deity (e.g. scheme) from another, let alone their mutual incommensurability. Much more usual has been to see them as big spongy things: networks, always evolving, by reconnecting and interconnecting. Not sealed off from each other.

Which is fragile enough as a positive creed, since conceived as the large-scale composition of myriad occasions of reference with no factual basis. So a reification of hot air (and ink). And seldom explicitly espoused, even by believers in the various big spongy things. (But go Quine!)

As a believer (in Quine's web for example), I should like to read Davidson as just saying to the extremists, the incommensurabilists: look, accept that if two links in 'separate' referential webs are in some undeniable way equivalent, then the webs can't be as separate as you thought.

But I have to admit he ends up happy to seem atheist about the "very idea". Is he? Does he mock the faithful?

---------------------------------------------------------------------------

A table of terms that are associated with both organize and fit — Moliere

I read it as pretty much Piaget's contrast of assimilate and accommodate.

OK, I'll try again.

One of Davidson's conceptual scheme possibilities is the Fit-Reality combination, right? He argues that alternative conceptual schemes in this case would have to be largely 'true' but not translatable. He then argues that 'true' and translatable are the same, the truth of some statement in the language of one scheme is simply it's translation into another - because both schemes are talking about the same beliefs, if they weren't we'd simply better presume we got the translation wrong.

But in this particular arrangement of a conceptual scheme, we're not talking about fitting experience (that would be another of the four), we're talking about fitting reality.

Nothing problematic for Davidson so far, because fitting reality is no different from being true. Being true doesn't add anything to the expression of what is.

This is where the paper comes in. Something can be 'true' if the scheme's predictions match the phenomenal experience, but the phenomenal experience is mediated by the scheme - we see what we want/expect to see, we literally make it true (now in the Ramseyan sense - which is the only way we can use 'true' here because Tarski doesn't apply). So for Davidson's fourth possible idea of what a conceptual scheme could be (Fit-Reality) the truth-theory of the language that conceptual scheme has isn't enough to say that no two schemes are 'true' and yet not translatable.

One person's scheme might include relations between objects which entirely match their phenomenal experience, precisely because they're altering their phenomenal experience to make it match the relations they're expecting to experience. Those relations would be 'true' for them, entirely translatable as relations to us, yet not 'true' for us who have different phenomenal experiences because we're not expecting them to be 'true', because of our model which does not allow for them.

The thing is, all this happens at the level of perception, sensory input in general, which is where I think we need to accept the idea of different, possibly incommensurable, ones existing. It's very possible that Davidson would like to say that conceptual schemes are something different, but in allowing for a Fit-Reality option in his definition, he's opened up this possibility.

Any tertiary sources? — Banno

I think Wikipedia's article is basically OK, but I'm no expert on this, so I couldn't vet it.

TL;DR - applying Tarski to conceptual schemes assumes underlying beliefs are the same, the article demonstrates they're probably not.

@Banno

I should add, just to be clear - I'm not 100% sure that Davidson even does use Tarski to dispose of the Fit-Reality option. He definitely does with the Fit-Experience option, but one of my issues with the paper is that I can't really see where he deals with the Fit-Reality option at all. So this may well be where I've gone wrong - I'm attacking the wrong argument because I haven't seen/understood the argument Davidson actually makes against Fit-reality types.

I wrote a synopsis of Friston's general approach outlined in the linked paper here. In terms of how they relate to external events/the environmental stimulus, a rough picture is:

The external stimulus behaves in way $\gamma$ (say light frequencies reflecting off an object) which impress upon us in ways $\theta$ (say light frequencies and intensities we're visually sensitive to in the context of the environment). We have a causal model of our environment (broadly, "this (situation) comes from that (expectation of action effects)") that relates our actions $\alpha$ and these impressions $\theta$ into the total causal model $f(\alpha,\gamma)$. This represents how we interpret and act in an environment.

In terms of how we use this information, we map $\gamma$ to $\theta$ through bodily constraints. $\theta$ is what our mind/body uses as distinct information sources in our environment; they are environmental parameters (where stuff is, luminescence, light frequencies, item topographies for touch, heat sensitivity, bodily proprioception etc) as they impress upon us perceptually; as we perceive them. They filter our environment and condense information in it into actionable perceptual chunks.

We then map $\theta$ to environmental/action/self model samples that represent all possible information about $\theta$; this represents what perceptual features arise in our experience that are consistent with (and most probable in) our model of what our experience should be.

There are different "models" here, and the term is used loosely in the discussion so far. The relevant distinction I believe is one between $\gamma$ and $f(\alpha,\gamma)$, which is roughly $\gamma$ being "the causal structure of reality operating in our environment right now" to $f(\alpha,\gamma)$ being "our bodily capacities for being influenced by it insofar as they are perceptible" (consider that we react to alpha radiation on a cellular level, but can't sense its presence unaided).

Personally, I think that "The cat is on the mat" is indeed true when the cat is on the mat, and the model we have is an interaction with our environment that reveals some of its structure; in particular it can reveal that the cat is on the mat. It doesn't just reveal that "I perceived that the cat is on the mat", or that "The cat is on the mat with respect to my model" in usual circumstances our causal model of our environment is informative about how it works, it establishes all three (when in a context that perception is sufficiently reliable). The cat really is on the mat, it's also modelled as being on the mat, it's also perceived as being on the mat. It being modelled as being on the mat or perceived as being on the mat is not sufficient for "the cat is on the mat" being true. The only thing which is both necessary and sufficient for "the cat is on the mat" being true is for the cat to be on the mat.

In terms of the models, any individual's perceptions and actions are not causally separated from their environment; the way we parse an environment's causal structure has bodily constraints, and it need not faithfully represent any particular aspect of the environment's causal structure. We generally perceive environments in ways related to our concerns. We learn to see; so our perception is historically structured as well as bodily constrained as well as contextually informed.

In terms of "The cat is on the mat", I could say that after perceiving that the cat is on the mat. In terms of my perception, "The cat" is a perceptual feature; but perceptual features in usual circumstances are informative relations between environmental stimuli and action-perception chains (expectations and memories, protentions and retentions if you're feeling phenomenological about it); the perceptual feature is had by the perceiver, but it is nevertheless a relationship between perceiver and perceived. The presence of "the cat on the mat" as a perceptual feature is strong evidence (in usual circumstances) that there is indeed a cat on the mat, but the presence of the perceptual feature's evidentiary status with respect to the claim "the cat is on the mat" does not imply that "the cat is on the mat" is true if the perceiver sees it there. It's true only when there is a cat on the mat.

The models Friston talks about are like evidence accumulation machines given a particular set of expectations of how stuff works and what we can do; we perceive and act in order to minimise the difference between what we expect to happen and what is happening. Action tries to normalise the environment given a perception of its structure (in terms of perceptual features), perception tries to normalise proposed actions given expectations of environmental development (in terms of self modelling internal states, sensations etc). We store and are influenced by previous states on all levels; past actions influence future ones. Altogether, this paints a picture of us as a process of coming into environmental and bodily accord given goals and an environment and a body which we have partial access to and represent those accessed parts with some errors.

On my reading, this requires a distinction between what we expect to happen and what is happening, even if that distinction itself is something the model has purchase on; we adapt to minimise to these discrepancies. The discrepancies don't just come from our models, they come from our environments not being in accord with our models insofar as we are sensitive to the environment and our body.

One way this might interface with the current debate is that the model we have adds nothing to the truth conditions of "the cat is on the mat", Friston's account is a highly sophisticated rendering of what it means for perception to be embodied and active and model based and how this might operate neurally; that is, Friston's account spells out a scientific theory of how our perceptions and actions are theory ladened (even down to the level of perceptual features). I don't think it problematise the notion of the truth conditions of statements at all.

@Isaac and I had a similar "realist vs anti-realist" (though I think we're both different flavours of realist, really. I suspect Isaac of some kind of hidden anti realism, Isaac suspects me of some kind of hidden naive realism, was my take) discussion regarding Friston's work in that thread.

That's just your own conclusion about the same subject matter. — Isaac

You're right. I shouldn't have said anything.

We saw that any conceptual scheme worthy of the title must be true. — Banno

Rereading Davidson and the thread.

Can you connect this statement to a quote in Davidson? Where does Davidson measure the titular "worthiness" of conceptual schemes?

Davidson argues that to fit is to be true. — Banno

I see the argument here as a rejection of defining truth in terms of fit. — Banno

A contradiction here vis-a-vis truth in connection to fit. Can you parse this or connect it to some quotes?

I see the argument here as a rejection of defining truth in terms of fit. Davidson takes truth as fundamental. — Banno

Davidson presents a compelling argument that if a scheme fits the totality of sensory evidence it can be said to be true.

So this is one way to arrive at the truth of a scheme.

But he immediately rejects the notion of fitting. He says fitting adds nothing to the notion of being true.

Okay. Maybe it adds nothing. But "fitting" wasn't playing the role of "adding something to being true." It was playing the role of allowing language to say such and such a scheme is true. Indeed without the fitting there is no "being-true" to add something to.

A would call this moment in the essay (and in your exegesis) a "deflationary leap."

Can you explain this a little better?

@banno

The above is why I can agree the T-sentence makes no reference to a fact, but have to insist that without a fact the T-sentence can't be put to use to ferret out truth. It's just a nice formula with no content.

T-sentences — Banno

Next: "all the sensory evidence qua facts" is what makes a scheme true. Here Davidson assumes the sensory evidence qua facts can be accessed directly without the mediation of a scheme. Another assumption (no scheme mediates) and another deflation (from "fitting" to "facts").

Next: a deflation excising the notion of a fact. So the totality of sensory evidence provides us with the facts. But it's better not to think of facts. Just think of T-sentences.

So: from a scheme fitting the sensory evidence (facts) to a schemeless access to facts to the excisement of the word "fact" in favor of T-sentences.

T-sentences — Banno

The deflation makes sense if it can be established that conceptual schemes share translatable content to a large (almost total) extent. Which raises the question of triviality and non-triviality in the distinction between conceptual schemes. (I mentioned this above.) The proportion of translatability to untranslatability has to be ascertained in order to justify Davidson's deflations, especially the deflation from a scheme fitting the totality of sensory evidence to (let's say) trans-schemic access to sensory evidence.

Can you explain this further?

The only thing which is both necessary and sufficient for "the cat is on the mat" being true is for the cat to be on the mat. — fdrake

Can you explain how this differs from the correspondence account (if you think it does) ?

... any conceptual scheme worthy of consideration will be true. — Banno

So those conceptual schemes deemed unworthy of the title will be those that fail to fit "the totality of sensory evidence" or those that contain some proportion or preponderance of untranslatability?

How do we access "the totality of sensory evidence"?

A digression, perhaps...

For Davidson, beliefs are caused.

...we literally make it true (now in the Ramseyan sense...) — Isaac

Ramsey's analysis of truth strikes me as very similar to T-sentences - instead of the simple T-sentence Ramsey had a process of translation from English to English* - an English without the predict "...is true".

As if "s is true IFF p" had English on the left and English* on the right.

So are you here talking about the Ramseyen sense of truth, or of belief...?

Because I suspect that if you were to replace truth by belief, modify this:

Something can be 'true' if the scheme's predictions match the phenomenal experience, but the phenomenal experience is mediated by the scheme - we see what we want/expect to see, we literally make it true — Isaac

to this:

Something can be believed if the predictions it implies match the phenomenal experience, but the phenomenal experience is mediated by the predictions - we see what we want/expect to see... — Isaac

you would have something not at all unlike Davidson's somewhat fuzzy notion of beliefs being caused.

So you may be explaining in detail much the same sort of notion of belief that Davidson espoused.

As the triviality of the case decreases the background of common beliefs decreases.

Accepting that some degree of translation is possible, the background of common beliefs will never decrease to 0%. In cases of increasing nontriviality, though the background of common beliefs may dwindle, it will never decrease to 0%. The limit-case-percentage is unknown.

In the example of Chinese and Western medicine, as Banno describes it, the background of common beliefs may be as limited as a belief in tumors and the will to rid the body of them.

Within the two conceptual schemes there may exist two distinct conceptions of the body, two distinct conceptions of the mind-body relationship and of the magical potencies of matter and of metaphysics in general. This is exemplary of a nontrivial case. But there is agreement within the framework of both schemes that tumors are bad and should go away and agreement that they sometimes do go away. — ZzzoneiroCosm

The distinction between trivial and non-trivial ideas within conceptual schemas is a nice insight and a most salient point to make. The agreement about the undesirability of tumours though, I would say is not really a feature of the conceptual schemas themselves, but a "meta" feature, which is due to the threat to well-being and even life that tumours present; a threat which is felt by all who wish to live without suffering. Of course healing schemas in general are designed to serve this basic human desire to be free from illness.

Does he mock the faithful? — bongo fury

I was going to write that I think he does -

, I should like to read Davidson as just saying to the extremists, the incommensurabilists: look, accept that if two links in 'separate' referential webs are in some undeniable way equivalent, then the webs can't be as separate as you thought. — bongo fury

But that's not a bad summation.

But that's not a bad summation. — Banno

"Not as separate as you thought" is a far cry from "killing relativism."

@fdrake, @Isaac

Drop the neuroscience for a bit. Lets' look at an analogous situation. Suppose that one made a Bayesian representation of the economy, such that one could predict economic futures with some degree of success.

Should we consider Davidson's article to apply to this model?

I think we would do so only if someone were to claim that we talk about was this model only; that talk of any other economic models was incommensurate with this model, and that we can never just talk about the economy, only economic models...

The sin Davidson is castigating is that of thinking we can not talk about dollars, but only about economic models of dollars.

Honey, vinegar and flies.

Honey, vinegar and flies. — Banno

Always the politician.

In-group and out-group rigidity. Rigid exclusivity of the "view-from-everywhere"; something I anticipated and Cassandraed about somewhere above.

Vinegar, in this case, is the insistence on an objection.

We saw that any conceptual scheme worthy of the title must be true.
— Banno

Rereading Davidson and the thread.

Can you connect this statement to a quote in Davidson? Where does Davidson measure the titular "worthiness" of conceptual schemes? — ZzzoneiroCosm

Our attempt to characterize languages or conceptual schemes in terms of the notion of fitting some entity has come down, then, to the simple thought that something is an acceptable conceptual scheme or theory if it is true.

Davidson argues that to fit is to be true.
— Banno

I see the argument here as a rejection of defining truth in terms of fit.
— Banno

A contradiction here vis-a-vis truth in connection to fit. Can you parse this or connect it to some quotes? — ZzzoneiroCosm

The first is not a definition of truth, so much as of fit in terms of truth. The second says there can be no definition of truth. Hence, no contradiction.

yes...?

@ZzzoneiroCosm, your improved understanding of the article might be better facilitated by your attempting to answer these questions yourself.

your improved understanding of the article might be better facilitated by your attempting to answer these questions yourself. — Banno

You might be right. That's what I've been doing. Feel free to leave a commentary.

Davidson:

The point is that for a theory to fit or face up to the totality of possible sensory evidence is for that theory to be true.

This phrase - the totality of possible sensory evidence - seems extreme to me. Almost meaningless.

Davidson:

The totality of sensory evidence is what we want provided it is all the evidence there is; and all the evidence there is is just what it takes to make our sentences or theories true. Nothing, however, no thing, makes sentences and theories true:


By "all the evidence there is" Davidson appears to signify: The totality of possible sensory evidence.

What is "the totality of possible sensory evidence" and how do we access it? Does that phrase makes sense to you? Please explain.

noticed that this is not Davidson view; but Davidson’s Account of his antagonists view?

The first is not a definition of truth, so much as of fit in terms of truth. The second says there can be no definition of truth. Hence, no contradiction. — Banno

The first says that to be true is to fit (presumably the facts). Why would you say that does not qualify as a definition of truth? Also, how would that be different than to say that to be true is to correspond to the facts? If it is no different, then why would you claim that the correspondence account of truth is not one you agree with?