The labels 'continental' and 'analytic' are silly but I find it helpful to think of there being arty philosophy and sciency philosophy. They can be mixed in the same sort of way that architecture and gardening mix science and art.

But when philosophy is not disciplined by semantics, it must be disciplined by
something else: syntax, logic, common sense, imaginary examples, the findings of other
disciplines (mathematics, physics, biology, psychology, history, …) or the aesthetic
evaluation of theories (elegance, simplicity, …). — Williamson

I'm mainly interested in philosophy that is disciplined by mathematics, physics, biology, psychology, history, … , AI, …

Philosophers who refuse to bother about semantics, on the grounds
that they want to study the non-linguistic world, not our talk about that world, resemble
astronomers who refuse to bother about the theory of telescopes, on the grounds that they
want to study the stars, not our observation of them. — Williamson

Philosophers must communicate, but they are not obliged to communicate in natural language. Mathematics is a language and in particular, probability theory and statistics provide a much more expressive language than logic. There are also programming languages. I wish more philosophers would learn these languages and use them alongside natural languages. I have Knuth's literate programming in mind as an exemplar.

I am very dubious about using natural language as a tool for reasoning or "using words to think with". That's a double-plus bad telescope.

Dummett’s requirement that assertibility be decidable forces assertibility-
conditional semantics to take a radically different form from that of truth-conditional
semantics. Anti-realists have simply failed to develop natural language semantics in that
form, or even to provide serious evidence that they could so develop it if they wanted to.
They proceed as if Imre Lakatos had never developed the concept of a degenerating
research programme. — Williamson

To my surprise I find myself feeling sorry for anti-realists: have they no competent proponents? I haven't thought much about realism versus anti-realism, and I don't care about the issue. But it's a puzzle, a challenge, to develop natural language semantics for an anti-realist position.

I'd start by thinking about programming an AI agent which learns 'everything' from scratch using
reinforcement learning. This kind of AI is the opposite of LLMs: instead of trying to cram as much human knowledge into a machine as possible you force the agent to work almost everything out for itself.

If you look at the diagram on Wikipedia you'll see there's an agent and an environment. It seems that we are to take the environment as existing independently from the agent. But I look at it from the point of view of the agent: there is state coming in and action going out, but how could you program the agent so that it was a realist even if you wanted to? And even 'there is state coming in and action going out' is saying too much too quickly, for how can the agent even distinguish coming in from going out? In order to construct semantics for an anti-realist position I'd start by answering this question. It's a long long route from there to a community of such agents which communicate using something like natural language but I believe it's possible.

My response: Those who jump too quickly to an answer to "what are things made of?" fall; not water, not fire. The doubters have it right: we can intelligibly ask what bread is made of, but not, at least amongst the presocratics, what everything is made of. It is a step too far to ask what things in general are made of. — Banno

Whether we like it or not, and whether we intend to or not, we cannot will ourselves to confine our method to the study of bread rather than the world in general without already presupposing as its condition of possibility a general and primordial origin, that which is always and for everyone the case, regardless of how relative, subjective and contingent the experience. Subjectivty , relativity and contingency only emerge as what they are due to this ‘ general’ and primordial origin.

A bit of history seems appropriate here, given the comments above. A potted history, leaving out whole continents of philosophical discourse in order to map a rough path.

End of the century before last (I have to get used to writing that), the dominant philosophical system was Hegelian, of a British persuasion. Moore and Russell reacted against it, Moore by showing how the way it was articulated was far removed from common sense, Russell by showing how it was far removed from the precision of the new logic developed by Frege.

I am very dubious about using natural language as a tool for reasoning or "using words to think with". — GrahamJ

Russell's student, Wittgenstein, adopted a similar line of thinking to yours, Graham, developing at least in outline a new language based on the new logic, that could set out all and only the true statements. Having solved philosophy, he went on to become a primary school teacher.

Meanwhile Tarski developed a description of Truth for formal languages that was correct - so far as it appleid to formal languages. Quine, over the puddle, took a different approach, examining how a natural language might be interpreted in a formal language, adopting a holistic approach.

Then Wittgenstein realised that being a primary school teacher was much harder than doing philosophy, and went back to Cambridge. Under the influence of a few very, very good thinkers, he realised that he hadn't actually been paying attention to the complexity and utility of languages. He did a re-think that radically changed the way that he approached the topic, by looking at how language is actually used.

(Meanwhile, a few of the chaps at Oxford did something similar, perhaps on hearing rumours of Wittgenstein's work.)

Nowadays few, if any, philosophers would consider replacing natural languages with a formal language. They don't need to.

Quine's student, Davidson, pulled much of Wittgenstein, Quine and Tarski together in a theory of translation that doubled as a theory of meaning, taking truth as a primitive, proposing that understanding a language involves grasping a theory that could generate all the true sentences in it. This is not a replacement for natural languages. It provides instead a translation of natural languages using formal tools.

It may be intuitive, , but it is also based on some very tight argument from Davidson and others, and formed the basis for much work at the end of last century.

Dummett famously argued that understanding a sentence should be tied not to its truth-conditions per se, but to a speaker’s ability to recognise or verify when the sentence is true. This position led him toward anti-realism: the idea that the truth of a statement is constrained by our capacity to know or verify it.

The realist/antirealist debate petered out in the first decade of this century. Part of the reason is Williamson's essay. The debate, as can be seen in the many threads on the topic in these fora, gets nowhere, does not progress.

The present state of play, so far as I can make out, has the philosophers working in these areas developing a variety of formal systems that are able to translate an ever-increasing range of the aspects of natural language. They pay for this by attaching themselves to the linguistics or computing department of universities, or to corporate entities such as NVIDEA.

Something like that.

Say the professor points to the board with a 2 written on it and says "That's a prime number"

A realist would say the professor referenced a state of the world. Davidson allows us to dispense with propositions and correspondence to understand this. But isn't Davidson's stuff offered as a possibility? There was never any empirical testing, was there?

An anti-realist would emphasize that meaning is use, and truth serves a social function. We don't need to get caught up in trying to understand what the professor is referencing. Reference is kind of poetic anyway. This view is also built on assumptions, and has never been "tested."

I think the conflict is really about two conceptions of the nature of thought.

he clearly doesn't believe that an exclusively language-oriented method is enough. — J

I differ with him here - philosophical problems are overwhelmingly the result of poor choice of wording; to the point where that's an alternative definition of philosophical problem. Once the plumbing of language is done, what is left might be physics or politics but not philosophy.

So in this regard I am somewhat at odds with Williamson.




But we might agree on a methodology, such that working out a suitable language in which to state the problem comes first, then we see if there is anything left over that looks like philosophy.

Whether we like it or not, and whether we intend to or not, we cannot will ourselves to confine our method to the study of bread rather than the world in general without already presupposing as its condition of possibility a general and primordial origin, — Joshs

I'm unconvinced. Mostly because I don't quite see what you mean. We might start with the brute fact of bread, presumably, and work from that. No need for Plato.

There's a lot to unpack in that, and the worst outcome here would be yet another realism/anti-realism thread. This would be a good topic for PM, if it involves just you and I, or a new thread if others are interested.

But we might agree on a methodology, such that working out a suitable language in which to state the problem comes first, then we see if there is anything left over that looks like philosophy. — Banno

Sure, that's reasonable. As you know, I think philosophical disagreement is all too often only a wrangle over terminology, which is probably similar to what you mean. But once there's a tentative agreement on terms, what's left over does look like philosophy to me, at least enough of the time to be worth pursuing. "Solve or dissolve" sounds good in theory, but it seems contrary to the way philosophy has been practiced over the ages. Granted, a strict linguistic approach has an answer to that: It's been wrongly practiced. But it's not clear to me whether that determination can be made on a linguistic/semantic basis alone.

A huge topic, obviously.

So, page eleven, and the core complaint:

Much contemporary analytic philosophy − not least on realism and truth − seems to be written in the tacit hope of discursively muddling through, uncontrolled by any clear methodological constraints. — p. 11

And then:

We who classify ourselves as ‘analytic philosophers’ tend to fall into the assumption that our allegiance automatically confers on us methodological virtue. According to the crude stereotypes, analytic philosophers use arguments while ‘continental’ philosophers do not. But within the analytic tradition many philosophers use arguments only to the extent that most ‘continental’ philosophers do: some kind of inferential movement is observable, but it lacks the clear articulation into premises and conclusion and the explicitness about the form of the inference that much good philosophy achieves. Again according to the stereotypes, analytic philosophers write clearly while ‘continental’ philosophers do not. But much work within the analytic tradition is obscure even when it is written in everyday words, short sentences and a relaxed, open-air spirit, because the structure of its claims is fudged where it really matters. — p.11

That "assumption that our allegiance automatically confers on us methodological virtue" is quite accurate. You can see it in the reply I made to @Joshs, a couple of posts up. I didn't spend much time on the reply at all, instead presuming that my lack of understanding was down to a lack of clarity on Joshs' part, and so I threw the post back at him, expecting him to do the work of clarification. Quite rude, by some standards.

Trouble is, I think that what I did is the right approach. It should be down to the poster to make their case. And I think Joshs would agree, but perhaps say that he had made his case sufficiently, and I should be able to follow it; that it is my lack of comprehension of certain philosophers from outside the analytic tradition that is at fault.

And it's not clear that we cannot both be right.

What is clear is that there is much more that needs to be said, were Joshs and I to pursue that discussion. And so to

...we should be open and explicit about the unclarity of the question and the inconclusiveness of our attempts to answer it, and our dissatisfaction with both should motivate attempts to improve our methods. — p. 12

but also

...it must be sensible for the bulk of our research effort to be concentrated in areas where our current methods make progress more likely. — p. 12

We ought pick our fights with care.

(, happy to come back to your point. I'm not saying that your point is not worthwhile - how could I, if I haven't grasped what it is?)

Russell's student, Wittgenstein, adopted a similar line of thinking to yours, Graham, developing at least in outline a new language based on the new logic, that could set out all and only the true statements. — Banno

No, no, no, no, no, that is not anything like my position.I don't know much about Wittgenstein, but enough to know I prefer the later version. (See my comment here https://thephilosophyforum.com/discussion/comment/985967 for example.)

I embrace Box's position 'all models are wrong but some are useful'. This is nearer to using a formal language in which only false statements will ever be made! Probability and statistics avoids the worst of the errors, which is why I want philosophers to use the language of probability and statistics.

The realist/antirealist debate petered out in the first decade of this century. Part of the reason is Williamson's essay. The debate, as can be seen in the many threads on the topic in these fora, gets nowhere, does not progress.

The present state of play, so far as I can make out, has the philosophers working in these areas developing a variety of formal systems that are able to translate an ever-increasing range of the aspects of natural language. They pay for this by attaching themselves to the linguistics or computing department of universities, or to corporate entities such as NVIDEA. — Banno

Thank you for that account. It sounds... not ideal, but a lot better than I had imagined. It's NVIDIA.

I'm aware I didn't respond to your comment in the "Belief as emption" thread, where you drew attention to some similarities with Bayesian stats. I recall thinking I would come back to it, but don't think I did.

Again, with these comments, I don't see a clear way to respond. I don't see why, for instance, "a formal language in which only false statements will ever be made" could be a "better" (more foundtaional?) tool for understanding reasoning than lambda calculus. It's not clear that formal logic and probabilistic reasoning are opposed. In fact, there’s a rich space of logics — Bayesian logic, probabilistic lambda calculus, epistemic logics with uncertainty — that treat probabilistic inference as a continuation of formal logical methods, not a rejection of them. So I wonder whether your contrast isn’t overstated.

I don't know how to explain, because I don't know how much you already know. In the other thread I mentioned the four components of Bayesian decision theory: model, prior, data, utility. Are you familiar with these? Could you put them to use in a simple example?

I am reading SEP's account of Bayesian epistemology. How you you get along with that?

Once the plumbing of language is done, what is left might be physics or politics but not philosophy. — Banno

And what does the honest philosopher (language plumber) think politics is? Total bullshit? Totally kidding ourselves about “natural” rights and platforms and ideology and voting for worldview change? All of these value anssignments and statements have no solid source, similar to plumbing. There is no true plumbing outside of the plumbing of language. Correct?

Amen, if that is how it has to be.

...we should be open and explicit about the unclarity of the question and the inconclusiveness of our attempts to answer it, and our dissatisfaction with both should motivate attempts to improve our methods. — p. 12

Yes, and notice how this paragraph begins: "Of course, we are often unable to answer an important philosophical question by rigorous argument, or even to formulate the question clearly. High standards then demand not that we should ignore the question, otherwise little progress would be made, but that . . ." and then your quote follows.

My worry about both (some) analytic phil and (some) Witt-derived phil is that the thus-far unanswered questions are indeed ignored, or rather ruled out as nonsensical. "Solve or dissolve," in other words. Let me ask you directly: Do you think there is a warrant for that, or is Williamson correct here? This clearly goes to the heart of the meta-discussion about method.

And what does the honest philosopher (language plumber) think politics is? Total bullshit? — Fire Ologist

The pairing of politics with physics suggests an answer. Neither is bullshit in the least, but (on this view) neither one is philosophy either.

And what does the honest philosopher (language plumber) think politics is? Total bullshit?
— Fire Ologist

The pairing of politics with physics suggests an answer. Neither is bullshit in the least, but (on this view) neither one is philosophy either. — J

Let me put the question more rigorously (I slipped into metaphorical speak feeling the use of “plumbing” gave me some leeway. I like the language plumbing metaphor by the way. Useful.).

My Question: If analysis of what is and is not done with language, or analysis of what can and cannot be done with language, is the most scientific of philosophical activity (so much so that one would say “once the plumbing of language is done, what is left [is] not philosophy”), then can any other discipline that relies on language be considered truly “scientific”? (“Total bullshit” was my metaphorical and absolutist way of saying “not science” or “not philosophy”). Or asked another way, if philosophy of language is the most scientifically rigorous use of language, are not all other uses of language less scientific?

By raising the pairing of physics and politics, it seems you are placing activities we do with language on a scale where maybe analytic philosophy (logic, Frege to Witt to etc.) is the most purely scientific pursuit, and then below that, something like physics admitting additional suppositions and assumptions to its game and way of thinking, history admitting more, politics admitting many more, metaphysics admitting more than that, maybe as much as poetry, myth and legend; and then fiction not even pretending to be science but is purely supposition. But language about language remains the clearest domain of the most scientific statements we can make. As in, “Once the plumbing of language is done, what is left might be physics or politics but not philosophy.”

Is this all in the right neighborhood of what Banno is saying?

The other notion that I might be missing here is that the measure of progress in science has emerged from sciences like physics, and not from analysis of language. We learned from physics how to be rigorous and how to measure progress, and then applied this as a tool to philosophy, and all that remained standing of “philosophy” as it was formerly called, after we applied this new tool, was analysis of language, the remaining philosophy qua philosophy.

But if that is the case, doesn’t that position philosophy, even if it rigorously keeps its attention on language, as less valuable and less universally applicable type of knowledge? That could be ok and could be the case but doesn’t philosophy become the study of the best ways to talk about physics? So if you learn rigor by doing physics well, there is really never any need to narrow your mind to thinking about language and logic in themselves; philosophy merely shows why two physicists understand each other. If they already understand each other, who cares about the philosophy of it all?

Maybe that is the neighborhood (domain) of philosophy? On my bullshit scale, physics is most scientific, and a sub-issue within physics that simply says when and why physics language is logical or, not nonsense, is philosophy which applies the rigor it gleaned from doing physics to the language physicists use when talking about doing physics. But philosophy qua philosophy, oversimplified, is the glossary, or plumbing, within the physics text book.

Further, if a philosopher tried to tell Einstein “you don’t know what you are saying.” Maybe that is even true, but for a physicist like Einstein to pose something nonsensical to the philosopher might just be the philosopher not able to follow Einstein’s meaning, as opposed to Einstein not being rigorous in his language. New rigor will always emerge ahead of the philosopher’s ability to codify it, axiomize it, and analyze the language this new rigor produces.

In which case I am back to thinking there can be a hidden philosophic rigor in any subject, such as metaphysics or politics. And the philosophy of language (analytic philosophy) is always secondary and post hoc, and things like physics, or metaphysics or even poetry, or politics are really only accidentally different, but each, like physics, could produce their own rigor.

So in the end, maybe Wittgenstein has eliminated philosophy as some sort of universal science of all sciences. That may be true. And we have moved philosophy inside each separate subject of study. So that we have physics, and then philosophy of physics (language rigor surrounding physics speak); or we have politics and philosophy of politics (language rigor surrounding politics, or political science).

But the difference between the subject of physics and the subject studied by a politician, from what I can say, is a metaphysical distinction. So I am back to wondering if there is a subject known as metaphysics that can be rigorously studied as well as the others. Or are we kidding ourselves that there is a real object called “the political”? Is that a figment like a platonic form, an ideal? Can one really be rigorous about politics? Or anything outside of language?



None of what I am saying or asking is meant to refute the article. Merely to understand it. I see myself as using implications to discern things prior and post the idea of true philosophic rigor, or the idea of philosophy as a progressing science with a distinct subject matter. I am not trying to recover metaphysics. Consequences be damned. I want to understand as much as I can.

I'm unconvinced. Mostly because I don't quite see what you mean. We might start with the brute fact of bread, presumably, and work from that. No need for Plato. — Banno

All brute facts about things in the world are subjective, relative and contingent. Even so, they rely on idealizations. Where does the conceptual category of bread come from, if not an abstractive construction? Whether I cite Husserl and talk about the fact that the moment to moment perception of a spatial object reveals continuously changing sense of phenomena which we idolize as ‘this self-identically persisting object’, or the sense of bread within a Wittgensteinian language game, one is dealing with the contingent and relative. To stabilize it for self-reflective analysis is to idealize it. The only aspect of experience which ‘escapes’ ( because it is presupposed by) the changeable and relative is the temporalizing process itself ( the synthetic structure of pat, present and future). This ‘general’ origin is at the same time utterly particular, because it is not itself outside of time.

@Banno Having read SEP's account of Bayesian epistemology, I think the entry on Philosophy of Statistics, especially section 4 would be a better start.

And what does the honest philosopher (language plumber) think politics is? Total bullshit? — Fire Ologist

Perhaps he thinks that "political philosophy" is an oxymoron.

Is this all in the right neighborhood of what Banno is saying? — Fire Ologist

@Banno will have to speak for himself. I don't think so. I looked back to try to see where you got "scientific" from and couldn't find it. Could you explain why you're casting this in terms of what is most or least "scientific"?

The Wittgensteinian Ur-picture, which I don't share, is that "philosophy leaves everything as it was." It is a diagnostic tool to help us understand where our language led us astray. Once we've done that, we'll be left with very little to worry about. Genuine problems will be assigned, or promoted, to the disciplines that study them, such as physics and politics. You can see why this is often viewed as a therapeutic understanding of philosophy -- or, less elevatedly, as plumbing out the pipes.

I think this is what Banno is describing. Again, he will tell us, I'm sure. Personally, I think a dose of Doctor Witt's therapy is a very good thing for all of us from time to time, especially when we get a strong hunch that our terminology is backing us into implausible corners. As I said to Banno above, I don't think all the important philosophical questions can be treated and dissolved in this way, but it's a fantastically useful technique to have at the ready.

Could you explain why you're casting this in terms of what is most or least "scientific"? — J

Dissecting, analytic, rigorous, scientific philosophy.

Discursive, narrative, not as scientific (or at all scientific) philosophy.

ADDED:

rejecting the suggestion that the mere divorce of sciences from philosophy is sufficient to explain progress — Banno

Instead of a sharp line between science and philosophy, place the analytic Witt type activity as the most scientific of philosophic activities, and the others fall below it. Isn’t the way I’ talking here in the spirit of the article? (Despite it not being rigorous, like the author admits he isn’t being rigorous.)

Personally, I think a dose of Doctor Witt's therapy is a very good thing for all of us from time to time, especially when we get a strong hunch that our terminology is backing us into implausible corners. As I said to Banno above, I don't think all the important philosophical questions can be treated and dissolved in this way, but it's a fantastically useful technique to have at the ready — J

Understanding Witt’s ‘therapeutic’ project in the context of consonant efforts in phenomenology and poststructuralism allows us to see that he doesn’t so much dissolve all philosophical questions as shows us that scientific , logical and mathematical domains are not self-grounding but instead are contingent and relative products dependent for their grounding on an underlying process of temporalization. Unlike writers like Husserl, Heidegger and Deleuze, Wittgenstein was reluctant to call the questioning that uncovers this process philosophical. He thought of philosophy as the imposing of metaphysical presuppositions (picture theories) on experience but not the self-reflexively transformative process of experiencing itself.

Section four is about the problem of the priors. That's a genuine difficulty for Bayesian approaches. However it is clear that there are some interesting developments in the area. I encourage you to start a new thread, maybe taking some of the novel results mentioned in Section Seven, and start a discussion about how a Bayesian approach might be helpful. In the process you might be able to show how you think it might overcome the limitations given in the article.

Have fun.

Genuine problems will be assigned, or promoted, to the disciplines that study them — J

That sounds good in theory, but how does it play out to the person interested in making progress? For instance, doesn’t that mean telling a politician that “liberty and equality for all” may be an incoherent fantasy and should probably not be discussed? Can’t really discuss non-existent universals and Platonist ideals, and impossible frameworks called “free agency” as if they can be demonstrably measured for all to see.

I’m just using politics as an example. I don’t care to discuss politics.

I’m saying, if the article successfully convinces the reader about the rigor that should be applied to all things spoken, (the rigor I have been calling scientific), then you can set up a game called politics and stay rigorous during game play, but once you question the existence of or value of the game itself, you are being metaphysical, and run out of tools and rigor and measures. In other words, politics becomes the bullshit it may actually be. Same as metaphysics. And really this would seem to be true of Physics. Which is how “physics” was once Euclid and Ptolemy (which are now mostly bullshit.). We may one day think Einstein was a joke, like the earth being the center of the universe is a joke.

We are stuck on a sliding scale of bullshit, except when we talk about language and logic and make these the subjects of our science - the ultimate game. Or in other words:
“Once the plumbing of language is done, what is left might be physics or politics but not philosophy.”
— Banno

My worry about both (some) analytic phil and (some) Witt-derived phil is that the thus-far unanswered questions are indeed ignored, or rather ruled out as nonsensical. "Solve or dissolve," in other words. Let me ask you directly: Do you think there is a warrant for that, or is Williamson correct here? This clearly goes to the heart of the meta-discussion about method. — J

Shouldn't we demand clarity as much from those asking questions as those seeking answers? So in Joshs' case, it is not just legitimate but incumbent to ask how we unpack "presupposing as its condition of possibility a general and primordial origin".

It's also legitimate, given our practical limitations, not to give full weight to every question, but to focus on those that appear most promising.

Analogously, not every reply in a thread deserves a response. It is at least to some extent incumbent on those posting to check their own work and see if they have erred, or could present a clearer case.

Why does the question remain unanswered? Why is it ignored?

Is this all in the right neighborhood of what Banno is saying? — Fire Ologist

No.

Isn’t the way I’ talking here in the spirit of the article? — Fire Ologist

Yes, I thought that might be what you meant, but since physics is science par excellence, I wasn't sure I understood you. Actually, it raises an interesting question: There is the rigor of science, such as seen in physics, but also something else in phil which doesn't claim to be science at all.

he doesn’t so much dissolve all philosophical questions as shows us that scientific , logical and mathematical domains are not self-grounding but instead are contingent and relative products dependent for their grounding on an underlying process of temporalization. Unlike writers like Husserl, Heidegger and Deleuze, Wittgenstein was reluctant to call the questioning that uncovers this process philosophical. He thought of philosophy as the imposing of metaphysical presuppositions (picture theories) on experience but not the self-reflexively transformative process of experiencing itself. — Joshs

All well said, thanks. The part I bolded is where the question of method, obviously, remains open for us. We need not agree with Witt about what constitutes philosophy, while still valuing his accomplishments, under whatever description.

Yes.

Shouldn't we demand clarity as much from those asking questions as those seeking answers? — Banno

Why does the question remain unanswered? Why is it ignored? — Banno

Yes, these are the right questions to pose. If you think they're legitimate in any given case, I'll take that to mean that you agree with Williamson to some extent. And yes, we can't address every problem, but must pick the most tractable and interesting.

All brute facts about things in the world are subjective, relative and contingent. — Joshs

Is this to be read as a stipulation? It doesn't correspond to, say, Searle's use of 'brute fact" as mind-independent, non-institutional and (at least usually) physical.

Even so, they rely on idealizations. — Joshs

Arguably, they are interpreted so as to be stated... We'd have to look in to what is involved in "idealisation" to see how that fits.

And so on, by way of sense-making - my putting in my terms what I think youa re saying, you putting in your terms what you think I am saying, such that we seek some common ground from which to see what is at issue. I'm sure you will agree that there is much more to be said here, and we could go on in kind for quite a bit. I think we'd be matching your more phenomenological approach to Davidson's triangulation, itself a huge topic, but one that might well be worth pursuing.

All somewhat tangential to the topic here, which is analytic method.

If you think they're legitimate in any given case, I'll take that to mean that you agree with Williamson to some extent. — J

Oh, yes. I think this the topic of the next few pages.