draft

The sage on the mountaintop (36 views of Mt. CritRat)

The critical rationalists want science to be like philosophy. That causes them to miss the difference between critique and experiment.

Link about this series

Sociologist Randall Collins provides the epigraph for this post: The Sociology of Philosophies: A Global Theory of Intellectual Change, 2009, p. 728.

In the field of fundamental issues which is philosophical turf, creativity is tightly focused conflict boring in on problems until deep faults are found; around these reconceptualization takes place. In this sense Popper accurately recognized in falsification something central to intellectual life – perhaps not in actual histories of scientific discovery, but in the world of the philosophers that surrounded him.

That’s the only substantive mention of Popper, which isn’t surprising: Collins draws his data from Chinese, European, Indian, Japanese, and American philosophy from ~800 BCE to ~1950 CE. As a minor philosopher, Popper can only get so much room, even in 1034 pages (excluding references and indices). That’s OK, because his focus is on philosophical schools that last generations. He wants to understand, in a sociological way, how such schools are started, compete with other schools, and eventually wither away (leaving space for new schools). I’m focused on what’s only part of a single school and, really, a single generation.

The Sociology of Philosophies has influenced me a lot, but I should say a bit more about the nature of schools.

As members of a social species, philosophers derive “emotional energy” from having their ideas accepted by philosophers they admire (members of their own school).
Schools are typically founded by individuals who have one or a few Big Ideas. Schools elaborate on those ideas, along the way developing school-specific approaches to solving problems. This collection of ideas, approaches, and a few other things is similar to Thomas Kuhn’s idea of a “paradigm” in his The Structure of Scientific Revolutions (1962). Kuhn does a famously poor job of explaining his concept – Margaret Masterman writes of it: “On my counting, he uses ‘paradigm’ in not less than twenty-one different senses in his [1962], possibly more, not less.” That’s from her “The Nature of a Paradigm,” Criticism and the Growth of Knowledge, 1970, p. 62. If you want to understand paradigms, I recommend her paper, which is on pages 59-89 of that volume, whose full text is available.
Schools establish themselves by becoming “the new hotness,” and attracting ambitious philosophers. A typical move for such a philosopher is to attack another school and show how some of its claims don’t stand up. The more fundamental the claim – the closer it is to the heart of the enemy philosophy – the greater the esteem in which the the victorious philosopher will be held by their fellows. (Social species, remember?)

In what follows, I’ll first make some comments about philosophical practice, then argue that philosophical habits of thought hurt critical rationalism.

Some topics in philosophy

Philosopher as legislator

Most intellectual fields started out as a philosophical topic. Medicine arguably forked early. Physics came considerably later – let’s say in the centuries surrounding 1600. Psychology was a part of philosophy until the late 1800s through early 1900s.

When a field splits off from philosophy, philosophy stops investigating the topic of that field, but it retains an interest in legislating what it means to do that field right. Hence there is a philosophy of mathematics, philosophy of medicine, and philosophy of science. The critical rationalists are in that tradition.

What do philosophers actually do?

Although “they search for capital-T Truth” is the easy answer, I like the one given by Gilles Deleuze and Félix Guattari, What Is Philosophy? They say that philosophers are in the business of generating new concepts. On page 2. A more thorough description from some secondary literature is “Philosophy is the creation or construction of concepts; a concept is an intensive multiplicity, inscribed on a plane of immanence, and peopled by ‘conceptual personae’ which operate the conceptual machinery.” Even the “for dummies” secondary literature on Deleuze and Guattari is difficult, so I’ll rely on the sound-bite description for this post.

I don’t think that’s all philosophers do, or that it’s what they think they’re doing, but this suffices for my purposes.

The nature of a concept

Plato founded a literal school, the Academy, as well as a metaphorical school, Platonism. Plato had a form of first-mover advantage in the number of concepts he introduced into the centuries-long philosophical conversation. “The safest general characterization of the European philosophical tradition is that it consists of a series of footnotes to Plato.” – Alfred North Whitehead. Perhaps the most important was his idea of the Ideal Forms. He argued that every physical giraffe is a reflection or instantiation of a literally real but non-physical Giraffe, which is the essence of, or template for, what it is to be a giraffe and not something else. I use giraffes in honor of Peter Adamson’s “History of Philosophy Without Any Gaps” podcast (recommended). He loves his giraffe examples. The Forms (or Ideas) live in some separate, timeless, unchanging world that lives “behind” or “above” our world. The world of Forms is not one we can access via perception, only by reason.

To a member of Plato’s school, a concept (Idea, Form) has a clear boundary. For any object X – be it a particular animal or a particular behavior – it is either in category Y (Giraffe, The Good) or it isn’t. When determining which, you can’t cheat by comparing an animal to a list of all the giraffes there are; instead, you must know, then evaluate, a set of properties such that (1) every Giraffe has all of them and (2) no non-Giraffe has all of them. There are complexities for this definition. See “Necessary and Sufficient Conditions: Problems for the Standard Theory.” I don’t know that they have any practical effect on the day-to-day work of philosophers.

However, you can’t define a concept with just any set of properties. Instead, the philosopher (or natural scientist) must find definitions that, as Plato wrote (channeling Socrates), “carves nature at its joints.“ Or, in the Jowett translation, “The second principle is that of division into species according to the natural formation, where the joint is, not breaking any part as a bad carver might.” This is, by the way, the dialogue where Socrates says writing was a mistake. Those who read will find “not truth, but only the semblance of truth; they will be hearers of many things and will have learned nothing; they will appear to be omniscient and will generally know nothing; they will be tiresome company, having the show of wisdom without the reality.” I feel seen. The idea is that bones are naturally distinct entities. Arbitrarily separating a bone into two parts violates its nature, as would be lumping two bones into one. You want to avoid the same problems when it comes to the concept Giraffe or the concept The Good.

Cue 2500 years of arguing about the right definition of The Good, Truth, Piety, etc.

Concept ambiguity and shift

Plato’s idea of Ideas has been challenged.

Ludwig Wittgenstein pointed out that “game” isn’t a well-formed concept in the platonic sense. Some games have winners and losers, but others don’t. Some games are exclusively played outdoors, but others aren’t. And so on. There is no single property that all games share. Instead, something is a game if it shares enough of the right properties with something that’s unquestionably a game. The common term is that concepts have “family resemblances.”

I picture a Platonic concept as a single circle that cleanly separates instances of the concept from non-members. A Wittgensteinian concept is a bunch of overlapping circles. Any point within any circle matches the concept.

Not all concepts have this built-in messiness, but an awful lot of the important ones do. For a related philosophical take, see essentially contested concepts. As an example of the difficulties of concepts, it might be fun to think about the differences (if any?) between the concept of family resemblances and that of essentially contested concepts.

A Platonist might argue that Wittgenstein is just arguing about how normal humans use words and think of concepts, which is irrelevant to the eternal, unchanging, and perfect World of Forms. But in the here and now, we have to use words to solve problems, which means we have to somehow deal with multiple overlapping meanings. For less philosophical and more cognitive science arguments than Wittgenstein’s, see George Lakoff, Women, Fire, and Dangerous Things: What Categories Reveal about the Mind, 1987, and Gregory Murphy, The Big Book of Concepts, 2002. They describe the modern view that concepts have fuzzy boundaries with some elements being much more prototypical or central than others.

So if you want to fairly evaluate someone else’s claim, you – annoyingly but unavoidably – have to spend some time checking that you understand what they mean by their words. It might be that a dictionary definition suffices, but you often should dig deeper. While dictionaries have definitions, they’re only attempting to be sufficient for ordinary conversation, not discussion in depth. People who call up dictionary definitions in technical arguments are – consciously or no – appealing to an outside authority’s moral force to end a conversation, not to resolve it. Intellectually honesty demands that of you when you think a claim is wrong. Honesty especially demands it when you think a claim is absurd. And it absolutely demands it when you think a claim is self-evidently absurd.

Such honesty gets more difficult because concepts evolve over time. What the Romans meant by “virtù” is different from what Machiavelli meant, which is different from our modern idea of “virtue.” They’re not separate ideas, but they’re also not the same idea.

So deeply analyzing a claim requires knowing what the claim-maker meant by their words at the time they wrote. This can be difficult, and it’s bound to be somewhat tentative. If unwilling to do the work, you could be explicit that you’re evaluating the claim against your own meanings. Or, if being tentative is not your style, you could just not voice your opinion. That’s possible, you know?

This form of philosophy was quite the rage during the time Popper and Lakatos were most active, but is no longer an awfully successful school.

Thought experiments

Philosophy makes heavy use of thought experiments. Some famous examples:

Ibn Sina (Avicenna) asks us to imagine a newly created human floating in air, eyes closed, nothing to hear, no parts of the body touching other parts – in a perfect version of modern-day sensory deprivation tank. Would such a person be aware of himself? Ibn Sina says yes, and from this argues for the existence of a soul.
John Rawls asks us to imagine we’re planning a society. We’ll be born into that society, but we have no idea whether we’ll be born rich or poor, sick or well, etc. From this, he builds A Theory of Justice.
John Searle asks us to imagine being in a room with a slot in the door. Slips of paper covered with weird squiggles are passed through it. Our job is to grab such slips of paper, look the squiggles up in a rule book to get new squiggles, write the new squiggles on paper, and pass that outside through the slot. Unbeknownst to us, we’re having a written conversation in Chinese. Searle asks: do we know Chinese?
Philippa Foot asks us to imagine a trolley that will by default kill five people but, if diverted to a new track, will only kill one. Is it moral to switch the trolley to the less-fatal track?

Here are some differences between thought experiments and scientific experiments: I’m not saying that scientists don’t make use of thought experiments – famously, Einstein used them to great effect. I’ll use this section later to highlight what philosophers generally don’t do that scientists generally do do.

They’re usually used to introduce an argument and make it seem intuitively plausible. They are what Daniel Dennett calls “intuition pumps.”
They don’t have to be physically possible. None of my examples have ever actually happened. Only the trolley problem even could happen.
Philosophers have much tighter control over the (imagined) environment than, say, agricultural scientists do. In the Floating Man thought experiment, all of the environment has been defined away.
This tuning is used to limit the number of possible results. The trolley problem gives only two choices, when a real-world implementation would provide others. For example, a railroad workers' union points out that if you let the front wheels pass the point where the tracks meet (the switching point), then switch the back of the trolley to the new tracks, it will halt safely, saving everyone.

Observations

Thus far, the physical, observable world hasn’t appeared on stage. How do philosophers use observation in philosophy (as distinct from telling scientists how to use observations)?

Unfortunately, “observation” is one of those words with a meaning that’s hard to pin down. There’s talk in philosophy of “observables” and “unobservables” and where to draw the line between them. Since I’m not a fan of binary distinctions, I don’t much care.

But I am a fan of prototype theory, a cognitive science approach that says a concept X will have more-or-less X-like elements. For example, a 25-year-old heterosexual male who lives in a big city with a roommate, dates widely, and plays a lot of video games is a more central example of “bachelor” than is the pope or a 75-year-old widower.

For observations, the prototypical example is something you perceive with your unaided senses (particularly vision). Pointing at a praying mantis and saying, “That’s an insect,” is a good example of observing something and then reasoning about it. Looking at the praying mantis through a mirror or corrective lenses is a slightly less good example. Looking at an amoeba through a microscope or the moons of Jupiter through a telescope is a still worse example of observation because it raises questions of whether those instruments work as they appear to do. (Are those spots on the sun entities or optical illusions?) And what of the Homestake experiment that “observed” the core of the sun by burying 400,000 liters of cleaning fluid 1,478 meters below the earth and counting how many new radioactive argon atoms were in said cleaning fluid every few weeks (the number being between ten and one hundred)? That’s pretty far from a central example of observability. This observation is heavily theory-laden, as it relies on theories that (1) the sun produces light by hydrogen fusion, that (2) said fusion will – at a known rate – produce radioactive boron, which (3) decays to produce a neutrino (even less observable than amoebae), and (4) said neutrino is unlikely to react with other atoms in the sun as it makes its way to the sun’s surface (unlike photons, which keep hitting atoms, being absorbed, then being re-released in decay, reaching the surface only after 100,000 to 1,000,000 years, having been downgraded from high-frequency gamma rays to visible light), whereupon it arrives at the big tank, with (5) some chance of hitting the right isotope of a chlorine atom, which (6) will produce radioactive argon, which (7) …

Observations are used in philosophy as both sources of inspiration and examples. By “inspiration,” I mean something like this (fictional! but plausible) story:

In the early 1950s, philosopher J.L. Austin was attending a wedding. He heard the pastor say, “I now pronounce you man and wife,” and thought, “Wait – that’s weird. What he said isn’t the kind of truth-valued proposition – “The cat is on the mat” – that we philosophers talk about. That reverend actually changed the world by speaking. He made something that wasn’t true become true.“ In many jurisdictions a wedding legally comes into being only after the pastor signs some paperwork, but in the social world of the wedding guests, it’s the speaking of the words that creates the marriage. After the reception, Austin retired to his study, built a whole theory based on such “performative utterances,” wrote How to Do Things with Words, and became one of the founding members of “ordinary language philosophy.”

That’s the use of observation in philosophy. What about as examples? Well, the neutrino example comes from a philosophy paper, Dudley Shapere, “The concept of observation in philosophy and science,” Philosophy of Science, 49 (1982) pp. 485-525. and the trustworthiness of telescopes (and mirrors!) is a core example in Paul Feyerabend’s Against Method. So real world examples are used in philosophical texts to explain or motivate ideas. But the text’s claims have to justified by more rigorous, step-by-step reasoning. Examples are more decorative than essential.

When it comes to observations, my impression is they tend to be comparatively small in number. Contrast Darwin:

Darwin was to spend eight hard years on barnacles, toiling away dissecting these tiny creatures under the microscope. The result was a 4 volume monograph on the Cirrepedia, living and extinct – the authoritative text on barnacles then and probably still now. “Darwin’s Barnacles”

This isn’t to say that such cataloguing is unknown in philosophy after the natural sciences branched off from it. The philosopher Mark Johnson (with his linguist coauther George Lakoff) collected a large number of examples for their Metaphors We Live By, and the Oxford natural language philosophers compiled lists of performative utterances.

But my sense is that such catalogs are smaller and less valued in philosophy than in the natural sciences. Biologists valued Darwin’s work on barnacles both because it was an exceptional achievement at the sort of thing they themselves did, and because it provided a dataset they could use for their own purposes. (Indeed, my understanding is that Darwin’s work on barnacles was a cornerstone of his reputation. Other biologists knew he sweated the details, which added credibility to his later theory of natural selection.)

Applying these intellectual habits to science

Here’s the second part of the Collins quote that started this post:

In this sense Popper accurately recognized in falsification something central to intellectual life – perhaps not in actual histories of scientific discovery, but in the world of the philosophers that surrounded him.

Popper and Lakatos were members of what Ludwig Fleck called a thought collective, “a shared framework of cultural customs and knowledge acquisition.“ Ludwig Fleck, Genesis and Development of a Scientific Fact, 1935. The quoted text is from Wikipedia. MENTION KUHN and PARADIGM? People in a thought collective don’t approach every problem afresh; instead, they view the world in the way people like them do. I argue that the philosophical thought collective caused the critical rationalists to overlay the philosophical habits I described above onto the practice of science. Consequently, their rules (I’d call them heuristics), while roughly useful, have big blind spots.