A narrow introduction, not a broad overview
The Eggplant is a narrow introduction to meta-rationality, not a broad overview. An overview would cover most or all meta-rational topics briefly, on their own terms. I hope to write that someday.
The aim here instead is to lead gradually from rationality into meta-rationality via a single narrow path: what would “true belief” mean?
We are not going to look in detail at the main concern of meta-rationality, which is “How can I figure out what rational methods will be effective in this particular context?” And, we will also not address the main concerns of rationality, “What should I believe?” and “How should I make decisions?” These are all important, and meta-rationality has much to say about them, but the questions “what are truth and belief anyway?” are prior.
My hope is to lead the reader to a beachhead: a landing place within meta-rationality from which broader exploration becomes possible. My aim is to give some readers a partial understanding, and to inspire curiosity in others.
The discussion only points out some major landmarks along the path to meta-rationality. It’s not a detailed guide. For some readers, it may seem unhelpfully vague, and inadequately supported by argument or evidence. The aim, though is not to convince anyone of anything. Rather, to sketch a map that might inspire you to find your own way through the jungle between the major waypoints, or to go looking for more detailed maps elsewhere.
Ready to move beyond STEM rationalism?
The ability to think and act using formal rationality is precious, and far too rare. I worry that rational thought and action are waning, due to current cultural dysfunction. Promoting rationality, and defending it against irrationalism, is urgent and important.
People differ in cognitive style, personality, and preferences. Some master formal systems, use them effectively for years, but eventually begin to find them somewhat limited and limiting and dry. Others continue to enjoy working inside a formal system indefinitely. If that is you, you may not find this book to your liking. Nevertheless, I commend you: you are keeping the world running, for everyone else, in the face of mass idiocy, hysteric delusion, and tribal selfishness. Please continue!
The Eggplant is likely to be mainly useful to people who have already started to recognize the limitations of rational systems, and who are increasingly curious about how to go beyond them.
I wrote this particularly for people with a strong STEM background. (“Science, technology, engineering, and math.”) It uses mainly STEM examples, and occasionally gets quite technical. However, most of it can be understood without any particular STEM knowledge.
It does expect a strong background in some discipline of rationality—business management or law, for example, if not STEM. I could have written it in terms of the limitations of administrative rationality, using transformational business case studies; but STEM is my first love.
A book in two parts
The Eggplant has two main parts: one on rationalism, and the other on meta-rationalism.
The first part works through a series of increasingly sophisticated rational models that try to explain what “believing true things” would mean. Each has a fatal flaw. Each can be patched, producing a more complex model, which also fails. Eventually it becomes plausible that no similar model can work. That suggests we need some quite different story.
The second part of the book presents the meta-rational alternative: a different account of how rationality works and how best to use it.
The book is long enough, and difficult enough, that it needs an exponential spiral organization—like the shell of a nautilus—with repeated summaries and expansions. So here is a more detailed overview…
Introducing an introduction
Before the first part of the book, there’s an introduction. It includes the outline you are reading now, and then two bits that clarify terms.
The first bit distinguishes reasonableness, rationality, and rationalism. There is an everyday usage in which “rational” means just “not stupid or crazy.” To avoid confusion, I call this “reasonableness.” Then there are many types of formal rationality, such as those found in science, mathematics, and law. These rational systems should be distinguished from “rationalism,” a body of mistaken claims about how and when and why rationality works.
The second bit introduces “ontological nebulosity”: the inherent indefiniteness of reality. Rejection of nebulosity is the root reason all rationalist theories have failed. Meta-rationality accepts nebulosity, and works with it effectively. Meta-rationalism is an explanation of how and when and why reasonableness, formal rationality, and meta-rationality work.
Taking rationalism seriously
Nowadays, rationalism often operates as a mere eternalistic belief system—an unfounded certainty based on imaginary understanding. Historically, though, rationalism was a serious intellectual project: to justify rationality by applying it to itself. The goal was a well-defined, detailed, rational explanation of what rationality is and why it works. Initially, there was no obvious reason this should not have succeeded.
The first half of The Eggplant works through a series of attempts, and shows how each failed. The aim is not to refute these theories in detail—because it is uncontroversial that each did fail. Rather, we will examine each failure mode in enough detail to diagnose the problem. By comparing these individual diagnoses, a pattern emerges: the overall reason rationalism can’t work.
On our “narrow path,” we look only at the question: “What it would mean to believe a true fact?”
- What sort of a thing would a belief have to be?
- How would the world have to be in order for beliefs to be true or false?
- What would the inside of your head look like if you had beliefs?
Each rationalist model fails to explain truth and belief for particular reasons. Each of these failures can be understood rationally, and so we can apply a patch to explain each class of anomalies, yielding a more complex and sophisticated model. (A “patch” is extra stuff added to software to fix a problem.)
Eventually, rationalism comes to look epicyclic: a hairy, buggy kludge of legacy machinery that still can’t explain key observations. At that point, one may simply hope someone can somehow make it work someday. However, analysis of its failure modes, followed by fundamental re-thinking, leads naturally to the quite different, meta-rational alternative.
The sequence of models roughly recapitulates the development of rationalist thought over the twentieth century. Particularly, it could be taken as a rough history of logical positivism, which was the last serious rationalist movement.
However, the goal here is not historical detail or accuracy, but understanding the intrinsic reasons each patch becomes necessary—as the failure modes of successive models come into view.
The final diagnosis is that every rationalist account gets wrecked, in stormy seas of counterexamples, on the black reef of ontological nebulosity.
If that already seems obvious to you, you could skip over the first half of the book, to go on to read about the meta-rational alternative.
Our diagnosis was that rationalism assumes a fixed ontology, which collides with the reality of ontological nebulosity. Meta-rationality takes ontologies as malleable. Ontological remodeling—reworking categories, properties, and relationships—is a key meta-rational operation.
Meta-rationalism remodels the rationalist ontology of truth, belief, and rationality. So, to understand what sort of thing meta-rationalism is (a remodeling), we need first to understand what meta-rationality does (remodel).
That is quite a complex and difficult subject. The first chapter of part 2 gives an intuitive introduction, relying heavily on specific examples from the history of science. It ends with an extensive discussion of the successive ontological remodelings of the category “planet.”
Taking reasonableness seriously
Meta-rationalism’s explanation for how and when and why rationality works rests on an understanding of how and when and why mere reasonableness works.
Informal reasoning is adequate for most everyday tasks, and it it does not assume a fixed ontology. It usually deals with nebulosity effectively. How?
To find out, we need to take reasonableness seriously. We should not dismiss it—as rationalism often does—as irrelevant because it is irrational, or imagine that it is a crude, weak-sauce approximation to true rationality. It addresses issues formal rationality can’t and doesn’t.
We need to investigate reasonableness as an empirical phenomenon. I will review some major features and dynamics of reasonable activity that have been discovered through rigorous observation.
In summary, reasonableness works because it is context-dependent, purpose-laden, interactive, and tacit. The ways it uses language are effective for exactly the reason rationality considers ordinary language defective: nebulosity.
Formal rationality aims for the opposite qualities: context-independence, purpose-independence, detachment, and explicitness. In some cases, it gains huge leverage by translating a problem from its nebulous real-world specifics into an abstract, formal realm, with a fixed ontology. It solves the problem in that domain, and re-applies the formal solution to the real-world situation.
Formal rationality depends on reasonableness for two reasons.
First, it relies on reasonableness to translate between the nebulous real world and a clear-cut formal abstraction of it. In the simplest cases—sometimes encountered in real-world application of physics theories—that is a matter of simple approximation. The formalism applies directly, to within a fixed margin of error. In most cases, however, it is a much more complex, and nebulous, matter of interpretation and negotiation. These are non-rational, but reasonable, cognitive activities.
Second, perfect context-independence, purpose-independence, detachment, and explicitness can never be achieved—in principle, never mind in practice. Formal models cannot entirely meet these standards for “rationality,” even setting aside the interpretation issues and computational limitations. Rationality is, therefore, much more similar to “mere reasonableness” than rationalist ideology supposes. Further, in practice, formal reasoning is almost always intertwined with informal cognition.
After meta-rational remodeling, “truth” and “belief” appear as jumbles of disparate, nebulous phenomena, sharing only a vague family resemblance—but nevertheless important. This is why meta-rationalism isn’t “sometimes it’s good to believe false things,” or “intuition trumps logic,” or “truth is just a social construction,” or “whatever you take it be,” or any of those other anti-rational clichés.
Rationality, on the meta-rational view, is not the optimal method for discovering truths. It is a jumble of disparate methods of understanding, which work more or less well in different sorts of situations. That makes it no less valuable. And, taking this more realistic view of how and when and why rational methods work can help us apply them more effectively.
Taking meta-rationality seriously
Meta-rationality is reasoning about which rational methods to use, and how, in a specific situation. That can include remodeling systems by understanding how they relate to contexts and purposes.
Effective meta-rationality depends on meta-rationalism: an accurate understanding of how and when and why rational methods work.
The Eggplant is not a manual of meta-rationality. The last part of the book aims more to explain what meta-rationality is, and ways it manifests, than to show you how to do it.
Nevertheless, it does explain some common meta-rational operations, and points out patterns in how they go. I plan to discuss this in more detail in follow-on work. However, I hope even the sketch here may have some practical use.
Does any of this matter?
STEM seems not to work as well as it used to. Hugely more research and development effort yields fewer breakthrough discoveries and fewer significant new products. New ideas seem increasingly hard to find. Many sciences now face a “replication crisis”: most supposed knowledge in these fields turns out to have been false. Eroom’s Law shows that it keeps taking exponentially more R&D to generate new pharmaceuticals. Part of problem is doing rationality badly: failing to preregister experiment designs, or using lousy software development methods, for instance. I don’t think that’s the main thing.
Too much R&D is mechanical by-the-book crank-turning, within a fixed framework, without reflection on whether it makes any sense in context. STEM can be “bad” not because it’s wrong, but because it’s trivial or irrelevant. Only a meta-rational view can help with that.
For individuals in STEM fields, meta-rationality becomes particularly important as one moves from being a technical contributor into management roles. Done well, management is an inherently meta-rational activity: it is about selecting, modifying, and creating systems, in the face of nebulosity. Entrepreneurship is even more obviously meta-rational: you create a company out of nothing, with no rules to guide you, only a nebulous understanding of a business opportunity.
Taking a still broader view:
Our societies, cultures, and selves also seem not to work as well as they used to. In How Meaning Fell Apart, I explained why. In short, the modern world was built on a foundation of rationalism. When rationalism failed, modernity ended. We live now in “postmodernity,” resulting from abandoning rationality, universality, and coherence. Postmodernity is characterized by nihilistic malaise, atomization, and political dysfunction. It could result in a civilization-ending catastrophe.
In Sailing the Seas of Meaningness, I suggest that recovering rationality is an urgent antidote. However, attempting to reinstate modernity by force, on a foundation of rationalism, is infeasible (even if it were desirable). The failures of modernist rationalism are too obvious.
Meta-rationalism is an alternative, more accurate explanation for the value of rationality. I believe we can and should and will remodel society, culture, and ourselves on a meta-rational basis. That will deliver the benefits of rational modernity without its harms and errors.