I found the following quote from Frege on the Language goes on holiday blog and it struck as the perfect starting point for this essay which has been written for a while now:
“Frege (“Logic in Mathematics”): Definitions proper must be distinguished from elucidations [Erläuterungen]. In the first stages of any discipline we cannot avoid the use of ordinary words. But these words are, for the most part, not really appropriate for scientific purposes, because they are not precise enough and fluctuate in their use. Science needs technical terms that have precise and fixed Bedeutungen, and in order to come to an understanding about these Bedeutungen and exclude possible misunderstandings, we provide elucidations. Of course in so doing we have again to use ordinary words, and these may display defects similar to those which the elucidations are intended to remove. So it seems that we shall then have to provide further elucidations. Theoretically one will never really achieve one’s goal in this way. In practice, however, we do manage to come to an understanding about the Bedeutungen of words. Of course we have to be able to count on a meeting of minds, on others’ guessing what we have in mind.”
Duncan Richter’s commentary then follows:
“Frege’s problem is of a different kind [from Mill]. There is something wrong with what he wants. He sees the problems himself, but still, apparently, goes on wanting the same thing. So pointing out the problems won’t help at all. We might say he needs a kind of therapy, although this won’t be regular psycho-therapy.”
Well, I have been thinking about the need for exactly such a therapy and it must stem from an understading that Frege was wrong about the extent to which we can in practice determine the precise Bedeutungen of our terms. As I hope to show below, the infinite regress of elucidation intrudes on our every day thinking in many ways that make even relatively simple communication or understanding difficult (a never ending process of negotiation). Difficulties stemming from what I call below the impossibility of perfect reference are not a matter of some distant perifery of hypothetical paradoxes, they make themselves known as insurmountable obstacles in seemingly innocuous. Or in other words, it is Erläuterungen all the way down.
And this problem does not have an epistemological solution (even if we don’t have to go as far as Rorty in rejecting epistemology as a beneficial enterprise altogether). Our only course of action is acceptance and making peace with the fundamental indeterminacy of reference. The acknowledgment of the need to make peace is the therapeutic part because the alternative is dissolution into madness of circularity or arbitrary absolutism (which is a kind of madly willful blindness, in itself).
Halting Problem of Rationality
The original impetus for these notes was reading a recent review of Elezier Yudkowsky’s new book on Inadequate Equilibria by Scott Alexander. Yudkowsky and Alexander’s review seem to me an object study of what I’ve come to thinking about as the halting problem of rationality.
This problem has many formal kindred spirits in the form undecidability, computability (P=NP), etc. From everything we know, we should be exteremely skeptical of rationality to solve its own problems without any appeal to a sort of axiomatic arbiter (a Godelian ‘because I said so’, perhaps.)
Scott Alexander shows the infinite regression of the process of finding the final level at which to decide which perspective is valid (or even useful). Based on Yudkovsky’s book, he arbitrarily (or perhaps magically) uses two perspectives but they are clearly just points on a continuum which itself is on an infinite plane rather than just a neat straight line.
Now, Yudkowsky does not seem to be bothered by the infinity of it all. He uses a whole lot of Bayesian heuristics to build up a priors machine that spits out one good decision after another. Prior ex machina, if you will. And it’s not always good. That’s why Alexander calls the book’s core argument ‘theodicy’. And that’s how most rationalist epistemological arguments strike me. They are the same sort of hermeneutics performed on the Bayesian heuristic canon that biblical scholars engaged in with the Bible. Read the text and its understanding will reveal THE truth.
The impossibility of true hermeneutics
My arguement is that hermeneutics (in this sense) is impossible and always the wrong goal. What’s more it is very easy to mistake our heuristics for hermeneutics. In other words, it is almost an instinct to assume that the analytic instruments we use to handle the world around us for specific (if often implicit) purposes are isomorphic with the world. And the more successful the instrument, the more likely it is we will assume it reflects the actual ‘true’ and complete image of the world. So computers, have been hugely influential and successful in emulating (and enhancing) some previously difficult mental processes and therefore the world is made up of information and our minds are just computers. We can control so much of the world around us by manipulating chemical elements, and therefore everything we are is really chemistry and our goal in describing the human condition should be a transcription into chemical notation because only that is the language in which a true image of the world can be captured. We can describe a sentence with a transformational rule and therefore the true representation of language is a formal description. We can design precise logical proofs for truth conditions, and therefore all that a meaning of a word or a statement is, is its conditions for truth. We can describe the utility of an economic transation by its marginal value and therefore all that defines value is the margin. And so on.
Richard Rorty pretty much showed how this works in Philosophy and the Mirror of Nature and later on also showed how to deal with it through his ironist approach. But rationalists are too cool to read Rorty. Wittgenstein and Derrida saw the problem and instead of talking about it, they tried to reveal it through cryptic koans.
I’d like to go about this differently and offer an outline of what a proof might look like that there is no ultimate external referent available for adjudication of referential problems. I also show that this causes problems not just on the edges but all the time across all aspects of academic and daily life.
Outline of the theory of everything
Lets start with a key assumption from which everything else derives:
On the word ‘exist’
Now, the word exist obviously has multipe meanings. I’m obviously not saying that everything exists as an object in the world. So I’m perfectly happy with the statement ‘Unicorns don’t exist’. I’m using it in the most universal sense similar to the logical notation E. In this sense it is impossible for something I can refer to with the word ‘something’ or even think about not to exist. But I don’t have to have a word or a thought for something to exist. In facts, words make it seem as if everything existed as some kind of entity. But those words and thoughts themselves exist and so does the relationship between them and the things they refer to as well as my reference to that relationship and my reference to that reference. And so on ad infinitum. In fact, the very act of naming brings things into existence. Existence in this sense is a Parmenidian totality – it is not temporal. Everything includes past and future. It is not dimensional – if it turns out there are infinite parallel worlds, everything will still exist. Parallel words are also part of everything. And if it turns out there’s no such thing, everything will still exist. The parallel worlds will just exist as an idea that turned out not to have identifiable external correlate. Everything does not require finiteness nor infinity. Infinity is still everything. But even if it turns out that infinity is just a mathematical construct and the physical world is actually finite in the shape of some bizzare multidimensional space-time sphere, that’s still everything. When Wittgenstein said ‘That of which I cannot speak, I must stay silent’, he was alluding to the same concept of everything. If it can exist it does exist, if it cannot, it does not. Everything exists. Anything that does not exist does not exist. What it means that there is nothing outside of existence in the sense of x E everything. There is no such special mode of being as metaexistence – existence beyond existence, existence about existence. Now, this is not the proof, this is the Cartesian axiom abstracted – X exists, therefore X exists.
Impossibility of perfect reference
The key consequence of everything existing is the impossibility of perfect referrentiality. This presents a problem because our entire epistemology is built on the assumption of referentiality. If something exists, we can refer to it with a concept, word, label, or at least point at it. In other words, signifier vs signified. We cannot speak or think without relying the perfect applicability of this abstraction. And most of the time it sort of works. In ‘Pass the salt’, ‘pass’ refers to an action, ‘salt’ to an object, ‘the’ to a relationship between the object and our perception of it. The ‘sort of’ refers to the fact that even simple sign/meaning pairings get very complicated very fast. Semioticians have been dining out on this since at least Peirce. (But medieval logicians and Indian ones before that have also taking this complexity apart as far as it can be taken apart.)
But it stops even sort of working very soon when we get even close to any attempt at metareferentiality. Just look what sort of verbal gymnastics I had to go through to even hint at what I mean by a simple statement ‘everything exists’. The problem is that referentiality is not a passive fact outside of existence. Every act of reference creates a new relationship between the refered, referee, and reference (at its most oversimplified). And that’s something we can then go and refer to, thus creating an infinite regress, that’s not linear but exponential. Because any new act of reference creates not one but at least four potential things to refer to. 1. The act of reference itself, 2. the referee in the act of reference, 3. the referent as being referred to, 4. the signfifier being used for that reference. Most often we can multiply that by referring to other participants in the act of reference, the relationship of that act to prior acts and their relationship to this act. In short, it’s not a pretty picture.
Borges in his psychedelic ways showed how the quest for perfect reference falls on its face in his short story about the mapmakers trying to create an ever better map but making it more and more closely resemble reality until it became as big as the land it was representing. By the end of his story, it simply lay abandoned on the edge of town. But the mapmakers did not even come close to achieving perfection. Because in the perfect representation of the world, the map itself would have to be included as well. But then an even bigger map would have to be created to capture the map, the reality and their relationship, but then we’d need another even bigger map to capture the previous relationships. And so on. A perfect map is a physical impossibility. Even in an infinite universe, there’s not enough transfinity to hold it.
There’s nothing new about this. Zeno, Russell, Goedel, Turing, Mandelbrot are just the most famous of the names who dealt with this problem in one way or another in the formal realm of mathematics. And Rorty did it for philosophy – while of course all the major philosophers of the last 300 years had hints of it, as well. Hume, Kant, Hegel, Marx, Nietsche, Heidegger, Wittgenstein, Rorty, Feyerabend. Of the western ancients Parmenides. But of course, so called eastern philosophy is rife with this, as well.
Meaning without perfect reference
So what does this mean? Is meaning impossible? Can we not speak? No. Meaning is obviously possible. But not in the way it suggests itself to us. When we say something means something we are implying a perfect one-to-one mapping of symbol to entity. But this is a false implication. When I point at an object and say this is a ‘chair’, I have a feeling that I have thus exhaustively described that object. That I engaged in perfect reference. But because everything about that chair exists, not just it chairness, I have simply pointed to a whole complex of existence and the word ‘chair’ only describes one of its infinitely many dimensions. When I set the chair on fire, at what point does it stop being a chair? When does it start being a chair during the contruction process? When the tree’s cut down with the intention of making furniture? When the last bit of varnish dries? Or somewhere in between? Maybe when it takes on the recognizable shape of a chair or when it can start functioning as a chair. What if it is a modernist chair and I can only recognize it as such when somebody puts a label on it? What if it is a chair in a picture? The label chair can do a lot of this work but it is not a perfect reference that maps nicely onto a thing.
This is all kind of obvious, so obvious that we take it all in our stride in our everyday acts of reference. But it starts causing problems as soon as we try to pin it down in the assumption that if we only stop being everyday about our reference, we can easily identify the ‘real’ referent exactly in the way our usual every-day unthinking reference suggests we are doing already. Oh, we’re just being sloppy thinkers, taking quick shortcuts for convenience. But if we sacrifice some of that convenience, take a bit more time, we will be able to stop the infinite referential regress. There has to be an end to it. But there cannot be. Not within the system of reference itself. Every moment we take to try to nail down the reference, creates another referent for us to refer. It just never ends.
Infinite perfect reference is impossible in principle. And we cannot resolve this by stepping outside the system of reference as we can do with maths in Godel’s theorem. Because, we can only consider reference using referential tools. This is so crazy-making and frustrating that generations of great thinkers simple assumed that it cannot be so. But in fact, it cannot be otherwise. Or if you think, it can. Show me how! I’ve been wrong before. (Obviously the Augustinian God who is outside of time – and presumably reference, or Buddhist karma – the extinguishment of existence itself – are pretty good conceptual exits out of the worry but they don’t provide any usable heuristics for dealing with the paradoxes of reference within the referential model itself.)
Summary of the key consequence
In summary, there’s a paradoxical consequence of the theory of everything. Beause everything exists, perfect reference is impossible, and therefore nothing exists in the way our words and thoughts make it seem it does. Or in a pithier (but less accurate) heuristic I recommend to all philosophers and rationalists:
“Just because there’s a word for it, it does not mean it exists.”
Edge cases in our midst
So what? Who cares about some edge cases on the margins of infinity? We can just happily go on to use our ‘ordinary’ language and take care of the really important problems like designing more efficient energy storage.
If only it were that easy. But as the example with the chair showed, the problems of reference are all around us. They pop up all the time in daily conversation or in basic academic discourse. They are not just something people in the most abstract domains have to deal with on their darker days. They are something we all deal with everyday – all of us – from Socrates to the Macedonian swineherd.
Lets take energy storage. It is a perfect way of thinking about batteries or pumping up water on an incline. But is there really a thing called energy we are storing the same way we may be storing bags of dried beans in a cupboard? Is there even such a ‘thing’ as energy? Well, there’s a whole lot of maths used to describe the measurements in the physical world that make it easy to think about a lot of things in terms of energy. Not only can we think of the world that way, we can all of a sudden compare things like burning fire and the rubbing of hands and the running of horses, pile of coal and pile of dinosaurs, etc. But what is happening when we say X is releasing energy? Is the pumping of water up a hill the same thing as a burning fire? What is it that we’re describing with the math? It is certainly not a given that energy is always a useful concept. People say things like, because everything is energy, I don’t believe in God but in universal energy that connects us all. We may laugh at charlatans like Deepak Chopra, but what is the mathematics describing energy really referring to? Is that one example of perfect reference? There is one energy and one value of energy in the world? Further indivisible? The ultimate building block of our semantics?
No. Theory of everything does not claim that no reference is possible. Or even that it is impossible to have one perfect one to one relationship between a signifier and signified. Just that that sort of atomic reference is not very useful. I can agree with my fellow referees that henceforth ‘dog’ refers to Spot at 5pm on July 23, 2011 in my living room (with the rest of the infinite specification taken as read). But that will render the word completely useless. I will have to then come up with a new word to refer to Spot at 5.01pm or Spot who’s wondered into the garden. Or I may choose the much more sensible option of refering to the fuzzy and ever changing universe of dogness. That word will be imprecise and fuzzy but that will make it useful. We will have broad agreement and negotiate around the edges.
So I can equally say, the word energy refers only to a set of mathematical formulae. But then I severly constrain what I can do with it. Which (in the case of physics) maybe exactly what I want. But it is a solution that does not scale as every effort to come up with a precise language has demonstrated and even if it did, it would necessarily run into paradoxes predicted by the theory of everything.
What are some possible objections to the theory of everything? I can think of several.
- The premise is wrong. Everything does not exist. There is a mode of metaexistence (for instance, human consciousness or a state of nirvana) that will make it possible to know all.
- There’s no problem. We just need an alternative epistemology which does not rely on reference.
- So what if perfect reference is impossible. We just need to come up with simpler formulae that will describe more complex ones and build a perfect reference by proxy.
- How does this apply to the theory of everything? How can you say everything exists when by your definition you should not be able to make any statements like that?
- You made a logical mistake and it is indeed possible to have perfect reference even when everything exists.
Re 1: Many years ago I read about a Buddhist school of semantics that claimed that the meaning of anything is everything that it is not. And the way Buddha himself was able to confirm that something does not exists was by looking at everything and finding that nothing was it. (This was a long time ago and I’m probably mangling this but it will suffice for illustration.) So is it possible that we can achieve some alternate level of consciousness – perhaps even stepping outside the ‘karmic wheel’ on which everything turns and grasping the whole world non-referrentially as one or simply being aware of everything through a vastly expanding consciouness where the limits of infinity don’t apply. Every mystical tradition would have you believe that you can.
But even if you could (and why not), it wouldn’t solve any of the problems in the here and now. Maybe we should realign our goals and instead of striving for accumulating ever more referential possessions, seek this new alternative consciousness. Sure. But again, this does not solve the problem for this consciousness.
re 2. Well, if you can come up with an epistemology not based around some notion of reference I’d like to see it. Now, there are many philosophical approaches that take the very impossibility (or at least great difficulty) of perfect (or even very good) reference to heart and integrate it into its epistemological toolbox. Zen Koans are one example, floating signifiers of post-modernist semioticians are another. But these approaches don’t actually transcend referentiality. They merely break it and through that breakage reveal the boundaries that reference imposes on us. The best Zen masters such as Derrida in his postcards or Wittgenstein in his investigations do a great job.
But, again this only exarcebates the problem rather than resolve it. There is no bulshit filter on koans. I can just as easily remain clueless as englightened and I have no way of knowing which one I am. Most of the reference transcending statements are as likely as not interpretted as if they are referential and simply referring to something not yet seen. Well, that does not help anyone.
re 3. The whole point of reference is that it simplifies the world. Who cares about perfection. As long as we can come up with simple and beautiful mathematics to describe the complex world, we’ll be in good shape. I call this generative referentiality. And if it could get us out of the jam, it would be nice. But it fails on two counts.
Count 1: Assume you come up with a nice function to describe a chunk of the world. Now, if you plug it into a computer, it will eventually spit out a perfect image of that chunk of the world. But then you’ve created a new object that needs to be generated by another function, including that function itself. Now, you might think that you could Cantor your way out of this. Just map one to one until infinity – no problem if it seems that one set should be smaller than the other. Yes, but Cantor never worried about everything. Russell did and look where that got him.
Count 2: But even if we assumed that generative referentiality can solve this problem, it is still arguable that it actually does do the job we assume referentiality does. Look at Madelbrot’s set. It is a dead simple formula (albeit with complex numbers) that generates infinitudes of self-similar shapes when plotted in a 2D space. But does knowing the formula actually constitute knowing the set? Can we know the set without knowing the formula? Do we need to know both? We can certainly take the formula as the signifier of the whole complex thing. But then it would seem to be mostly doing a job of referring to something complicated and calling it Bob (or Madelbrot’s set) would be just as good. There is something magical about knowing the names of things but knowing the names is not knowing the things. Generative referentiality is extermely useful and we might say it provides the foundations of our current civilisation. But confusing it with perfect referentiality has caused a lot of problems.
re 4. How do the epistemological limitations of the theory of everything apply to the theory itself? This is a typical worry of any foundational epistemological theory that tries to encompass all of cognition. How do you deal with self-referentiality without running into a paradox? The strictures are even more severe on any theory that tries to deal with self-referentiality itself. The limits on perfect reference of course apply to anything I say just as much as anything else. However, there is a small reprieve for reference that does not try to do anything useful. The whole point of reference is that it allows us to grasp something external to us. And the hidden strength of reference (at least hidden from most mainstream logicians) is that it is profoundly simplifying. It only works because it ignores almost everything and only zooms in on what is most important. However, there is a kind of perfect reference that is profoundly useless except as a foundational axiom. And that is tautology.
I can in fact avoid all the problems with chairs, love, kings of France or anything referentiality struggles with if I just say they are exactly what they are. So instead of positing that X = a, I simply say X = X. I can thus refer to everything as being everything and be quite happy that that reference includes itself and everything that surrounds it. Just like I can say that a set of all sets is a set of truly all sets including itself. The problems start when I try to build a non-selfreferntial system out of this assumption. Because I can’t.
I would say that the foundation of the theory of everything is purely therapeutic. It points to some fundamental impossibilities of our system without saying ‘and for my next trick, I will now show you how to simply resolve it’. There is no next trick. However, I will try to outline some heuristics that can be used to get around this. Deconstruction is one such approach – Derrida’s horizons come to mind here (but not something I know a lot about.) But even very simple rationalist heuristics will do as long as we don’t assume that they are external to the limits on perfect reference.
re 5. It is possible that I made a mistake somewhere. In fact, I would not be surprised in the least if I did – this kind of thinking is hard and not my strong suit. But what remains is the empirical fact that perfect reference is nearly impossible. It is so hard that nobody has yet managed to crack it in any system capable of expressing something like language. Even algebra. I never quite managed to understand the details of Goedel’s proof but this is what I imagine he was after. But for him, undecidability was an internal problem for any system with an outside observer. But with everything there is no outside observer. (Or at least not any outside observer we have access to.)
Words-as-models heuristic and the halting problem
So what are we to do? Perfect reference is impossible but our language-thought processes behave as if all reference was perfect. Is there a way out? No, there is no way out. You cannot be out of everything but there’s a way of living with this limitation.
One simple heuristic I suggest is to think of anything we say or think as a model. Each word, sentence, concept. It is a model of the thing it refers to. Then we can then go on and live with the statistician’s dictum: All models are wrong, but some are useful.
Of course, the world does not need me for this. Those assumptions have been around for a long time. But what has been missing is the next step. Ok, so some models are useful, how do we know which ones? Can we come up with a universal procedure for determining usefulness of models? And here the analogy with the halting problem comes in.
Models are a type of (by definition) imperfect reference. So, if we could get a perfect procedure for identifying the utility of models, we could build out a model of the whole world just based on utility. But utility of models is itself a mode and, therefore, by its nature imperfect. Which means we cannot have a perfect external procedure for identifying utility. So, what can we have?
As always, we need to remind ourselves of the heuristic ‘just because there’s a word for it, does not mean it exists’. We have a notion and a word of utility but that does not mean that there is a nice monadic entity of utility floating around in the world that we can attach that word to. We can pretend there is (just like the utilitarians) but that is not going to help us avoid paradoxes and other odius conclusions (just like the utilitarians). We don’t know whether a model is useful until we have examined all of its aspects with respect to all aspects of reality. But that is no more possible than it is possible to examine all steps of an infinitely recursive algorithm. At best we can follow the line of steps as far as the eye can see and say, well, it seems like it will continue for a while. Let’s go get a sandwich.
But with utility, things are even more difficult because it is not intrinsically a point on a simple scale from less useful to more useful. To simplify dealing with utility, we may convert it into a unidimensional scale of ‘utils’ spanning from negative to positive infinity. But that only makes the calculations of utility themselves easier by pushing all the difficult work one step down the line. We still have to decide in every case how to map the utility we perceive onto that scale. And we also have to decide how to measure that mapping. So by committing to a simple scale we have simplified one part of the process but we didn’t solve any of the problems. We simply pushed them upstream to the foundational issues.
How do we halt the infinite regress if we don’t know whether there is an actual end to it? In practice, we already do the only thing we can do. We give up when it feels right. Or when we’re exhausted. Or when we’ve reached a point of some sort of equilibrium or conversely leverage. Our only sane option is to do what we’re doing and not pretend that we’ve cracked the halting problem. Pretend (with conviction to the point of self-delusion) that we’ve come to a decision because a decision at that point makes sense. Dance as if noone is watching and there’s an externally arbitrated rational reason for stopping. Or a common sense one. But those are just pragmatic, ad hoc (or as Rorty insists contingent) decisions. The assumptions of external rationality are therapeutic ones, not epistemological.
Dealing with imperfect reference through heuristics: rationalists, postmodernists and pragmatists
Now, given we know all of the above and assuming we want to be reasonably honest about acknowledging there’s a problem, how do we go about continuing to speak and reason referentially while knowing that the reference we are working with as real is actually impossible? The postmodernists have suggested provisional knowledge. And they’re not wrong. All knowledge has to be provisional. The rationalists have come up with the Bayesian ‘strong opinions weakly held’ and updating priors. And they’re not wrong. And the Pragmatists have come up with conflating epistemology with ethics. I like these the most.
But these are just the general slogans of intent. What is really interesting (and actually useful) are the heuristics developed by each of these traditions.
The rationalists assume (implicitly) that perfect reference is indeed possible but very hard. They have come up (as the scholastics – Western and Eastern – before them) with a number of heuristics in the form of logical fallacies that help point out some of the paradoxes. They sort of present them as if avoiding these fallacies would avoid all problems. But while they help avoiding a lot of problems, they don’t avoid all or even most and they also create new ones. But simply dismissing them because of this would be foolish.
The postmodernists, on the other hand, focus on the impossibility of perfect reference and emphasize the provisionality of knowledge. They have developed a lot of deconstructive techniques to direct the mind to the boundaries of possibility. They almost write poetry about the abundance of everything and the futility of its conquest (Feyerabend being one of the most eloquent here). But they tend to reject even some of the more useful heuristics and are very likely to drown in bulshit. The rationalists are prone to non-sense, as well, but I think the profound embarassment of the Sokal hoax is unique to the postmodernists. The rationalists just assume that the infinite regress can be halted if we put up enough barriers of logic in its way, but postmodernists are sometimes all too happy to see something rhyme and don’t care if it could be made reasonable sense of (albeit provisionally) with some simple rationalist heuristics.
Then, there are the pragmatists. They are closest to my heart and I think Rorty pretty much said everything that I ever wanted to say. They emhasize the contingency of knowledge on situation and social commitments. But unlike the postmodernists, they are happy to take provisional stances for something and do something specific with them. When James spoke about the importance of commitments to others as being the foundation of epistemology, he touched on something fundamental. I came up with the slogan ‘epistemology is ethics’ without knowing about James or the details of Rorty’s analysis but when I read Philosophy and the Mirror of Nature, I knew Rorty and I were soul mates.
But I think Rorty was too quick to dismiss epistemology. He rightly took it down a lot of pegs and showed the impossibility of an ultimate epistemological theory. But he did not give it enough credit in thinking through some of the impossible problems while asssuming they are merely very hard. His ‘liberal ironist’ stance in later essays is a good practical application of the core insight but again, it does not give enough room to the basic heuristics.
That makes it much easier for the traditional epistemologists and scientists to dismiss him as irrelevant. While in fact, he speaks to the very core of their enterprise. But it feels to them like he is taking away the very foundations on which all of their heuristics stand and somehow invalidates them along with it.
But Rorty should be viewed as therapeutic. If I can hope to add anything to Rorty, it is this. Similar to the New Wittgenstein studies. Everytime we run into a referential paradox, we can take solace in its totality and turn away from the brink. We can also just simply save time and not worry about justifying stopping following the referential regress. But we can also let ourselves an out by remembering that we stopped simply for pragmatic reasons. And if new reasons (contingencies) appear, we can resume our journey along the infinite refrential web.
Serenity through disciplined conversation
What I am ironically calling ‘theory of everything’ is designed to do just that. Acknowledge that there is a problem and that there’s nothing that can be done about it.
Very much like Alcoholics Anonymous. The difference is that the wisdom to tell the difference between things we can and cannot do something about is not revealed by a deity but is a constant subject of disciplined conversation. Conversation that reflects the contingencies of the present as much as those of the past. A conversation that cannot have an end but which we must inevitably take part in. The serenity one hopes to get out of this will not come from resignation but from embracing of the totality without assuming that we can grasp its every possible aspect.
This is the therapy Frege needs. As do we all.