The following is a nonexhaustive nonrigorous list of philosophical 'takes', beliefs, or interests I subscribe to and would like to ramble about, in no particular order. Note that I'm not in academia or anything, and I surely have lots of silly misunderstandings. My intention here is not really to identify myself with a set of labels in any particularly strict way, but just to record my current situation of mental constructions and credences and whatnot.
I may or may not update this post as I update my mind. Some sections may later be spun off into their own posts. (I'm honestly not sure it was a good idea to lump them all together like this...)
Naturalism
...Or "materialism" if you like. This is to say that I do not believe in "spooky" things, like souls or gods of religions or so forth- there is no "supernatural", though what exactly counts seems not well defined. If there were ghosts, and they became evident to me in a way I couldn't deny, then I would believe in ghosts... But would I reject naturalism? I don't think so... Ghosts would simply enter into what I consider "nature". What would rejecting naturalism entail, exactly? I'm not actually sure, so maybe this is a meaningless label.
To say something more concrete, I think that the stuff of minds is not a fundamentally different stuff than that of the material world. That's the usual subject of 'supernaturalism', I think: assigning a unique substance to consciousness or some such perception-level object (dualism, idealism, various religions, etc etc). I reject all this; I am physics and so is everything else.
The "reality" of a priori truths is something I am not sure what to make of, though. Surely \(1+2=3\) or \(\text{not true} = \text{false}\) or \(y = r \sin\theta\) are not real or true in the same way as "It's cold outside today" or \(H\psi = i\partial_t\psi\). But surely the former statements are real and true, or rather, they really truly follow from certain axioms. I don't know. Do I need to read Kant?
Structural Realism
Metaphysics blind to physics cannot hope to come before it. But physics has this peculiar tendency to totally revolutionize its own fundamental ontology with each major breakthrough, so it would be naive to simply grab the structure of current physics as the basis of a metaphysics... And anyway, those non-fundamental theories are still quite successful, so it'd be nice to account for them, yes? The solution is structural realism- each theory captures something of the structure of reality, if not its essence or whatever. Classical mechanics is not fundamentally correct, but it nonetheless captures a pattern or structure which is really existing in the world (if only approximately).
Physics uses the word "emergence" to capture this structural relation, orienting theories with respect to one another. Emergent theories are limiting cases which appear under certain conditions, or when applying certain approximations (associated with but not merely differences in "scale"). Classical mechanics emerges from quantum mechanics, galilean relativity from special relativity, thermodynamics from aggregate particle motion, and so on. Outside physics, the word "emergence" seems to have been widely misused and abused in various ways, so I often avoid it.
The more revolutionary move, then, is to drop out the fundamental "essence" or substance altogether and claim reality as nothing but structure. There is reality, isn't there? There is a difference between map and territory, some Nature at the bottom of theory, yes? But I am inclined to think that raw reality has "nothing to it" other than its structure. Or at least, whatever "something else" it has is wholly irrelevant to anything we could perceive and measure.
The super-duper revolutionary next move is to drop out "objects" from the "objects and relations" couple. If there is nothing but structure, then (depending on what is meant by structure) there is nothing but relational patterns, and the 'objects' which hold those relations are irrelevant at least, nonexistent at most. This is obviously unintuitive, and is regularly accused of nonsense. But evolving beyond intuitions is one of the centrally compelling aspects of all this to me, anyway. If we're too scared to wander past our intuitions, will we ever really learn anything? Why should fundamental reality be intuitive to us, anyhow?
I am interested in category theory, and its relation to all this. The spirit of category theory is to reject internality for externality, blackboxing "objects" to just study their interrelations (morphisms). Category theory working as a foundations of mathematics (alternative to set theory) is taken to imply this is the 'right move', somehow, that maybe relations really do come before the objects they relate. Though to me it looks like the better insight is not that category theory's success implies externality should be primary, but that its interchangeability with set theory implies an interchangeability of internality and externality... But, uhh, I don't know. Maybe you could argue set theory is already sneakily externality-first because of elements only having functional relevance through their equality relations to other elements. Or something like that.
(Structural realism turned out to be a pretty strong throughline across this post, I think. It is apparently the backbone of a lot of my thinking. Can you spot how it factors into all the following discussions?)
Deleuze & Guattari
Deleuze's metaphysics is doing something very similar to what I just described, dropping out "identity" (~objects) from its fundamental position to place "difference" (~relation/structure) as primary. 'Things' are products in the more fundamental play of differences, which are the more basic stuff.
D&G then go on to do all sorts of interesting things I don't really understand. But the parallel to modern structural realism and such suggests to me they were 'ahead of the curve', in a certain sense, and modern metaphysics is only slowly catching up? Deleuze claimed somewhere his goal was a metaphysics which would accomodate modern science, which is exactly the origin the structural realism, and the results look similar (at least superficially). I'm currently reading "Every Thing Must Go", and it feels fairly Deleuzian so far.
Consciousness
I don't know what this view is called, but I see consciousness as not a fundamentally different stuff (as discussed before), but as a pattern or structure implemented by certain systems (also as discussed before). Philosophical zombies (p-zombies) are hypothetical people with no consciousness who nonetheless behave exactly as conscious people. There is an argument that the conceivability of p-zombies implies consciousness is something 'else', another stuff not accounted by physical people in themselves. This is silly to me, because the conceivability of p-zombies is something you are assuming...
The response which is compelling to me is that a hypothetical world full of p-zombies, since they act just as conscious people, would look no different from our world. They would write books and have arguments about consciousness and p-zombies, even, not acknowledging that they are p-zombies themselves any more than conscious people do. So the whole thing seems to fall apart. If p-zombies would believe the same things about consciousness that we do, then either p-zombies are impossible, or our beliefs are useless. For us to meaningfully discuss consciousness requires that consciousness be something which comes into play physically/causally, it cannot be something 'outside' which passively observes: If it makes no material difference, it cannot causally come into play in our beliefs, thus we have no grounds to believe in consciousness at all. But bothering fundamental physics to fit consciousness in somewhere is... well, terribly difficult at best, nonsense at worse. And it's a huge contrivance anyway, isn't it? Clearly there is consciousness, but clearly it is not some fundamental stuff; It is a pattern, a structure, built from physical stuff.
Free will and agency
I'll probably write more about this later. Agential concepts have been of particular interest to me in recent years.
It should by now be clear that I do not take seriously any 'spooky' free will, nor do I take fundamental physics's nonagential ontology as debunking the structural relevance of agency. I am aligned with some flavor of compatibilism. "Teleonomy" has been confusedly coined to refer to emergent teleology, but since there is no existing nonemergent teleology, there is no need for a concept of teleonomy.
A while ago I had a short period of skepticism toward agential concepts, due partly to my own executive dysfunctioning and foggy-headedness creating a real empirical rift between my sense of intention and action in daily life. I've since snapped out of this (not the dysfunction, but the rejection of agency). Often, I am thirsty, so I grab my cup, walk to the fridge, pour some ice into the cup, then some water, then lift it to my mouth and drink. This is a complex set of actions clearly set up to satisfy some preexisting desire. (If you think it isn't complex, try explicitly writing out ever micro-action, so a robot could do it.) It isn't post-hoc, because I could've written down my intentions beforehand. So clearly there is agency, it is clearly not all just an illusion.
But it is also clearly sometimes an illusion. We do things, and then feel retroactively that we were doing them in pursuit of this or that end, fooling ourselves. We see intentionality where there is none. One can argue that "free will" being apparently so baked into our intuitions implies good prediction power (which is surely the case to some degree)... But then, are we really understanding people because we have an actual working model for understanding, or because we are also people and thus can run a little empathic simulation and then retroactively frame the results as though they came from the model? (Anyway, free will intuitions vary by culture a good deal, don't they?) Social prediction power appears to come just as much from empirical precedent of individual actions (eg tradition) as from application of some integrated systematic agential framework, suggesting the agential predictions are really not good enough?
I am looking for a conceptual system which can ground agency in reality in a practical way. I don't feel like I've found one yet, nothing I've looked into has clicked in quite the way I would like. Dennett's "intentional stance" still feels hand-wavy to me. But, I don't know.
Gender stuff
This will probably get its own post, but I've basically entirely outgrown the views I expressed in older writings.
When children play tag on the playground, they designate one person to be "it". One could ask what material trait this person has, which defines them as 'it'? Of course, there is none- the identity label is in a certain sense arbitrary, it could just as well have been applied to anyone else. There is nothing particular about the person which makes them "it", no tangible "itness" about them. So surely this concept is meaningless, and we should place no importance on it, perhaps even do away with it altogether, right?
Then the game begins, and all the non-it children run away, while the designated "it" child runs after them and tries to tag them.
We are immediately confronted with the real tangible difference that "itness" makes, the influence it clearly has over everyone's behavior. There is a certain social relation between the children, a protocol or game they play amongst themselves: They agree that a certain person is "it", defined not by internal characteristics but by a shared understanding of how the role fits into a larger game, then they play along, acting within the roles that game defines. And they enjoy it.
Everyone is playing the gender game. (Though everyone disagrees what the rules are.)
Morality
You can probably guess that I am not a 'spooky' moral objectivist.
A part of me agrees "no ought from is" etc etc. But then, if oughts don't come from what is... well, that's just another way of saying they don't exist, right? There is clearly a reality to oughts in the same way I described the reality of gender. What I want to say is that morality is "real" insofar as it really operates in the world (through patterns of behavior), and not real insofar as it does not. Judgements are not a priori things with truth values ("this is bad"), but actions which take place physically in the world ("I condemn this").
Morality is usually discussed in the same way that statements of fact are discussed: As debate. It is presupposed that there is some 'fact of the matter' of moral judgements which one must be convinced into via reasoning... It seems true that, if you choose some oughts as axioms, you can work through some reasoning to derive other oughts which follow logically. But what axioms to use? To have a meaningul debate, you would need to start by sharing axioms, right? How often are people doing that? Are moral axioms even a common subject of thought among laypeople? It seems like a fringe academic thing...
Back to the operational picture, I think what is really going on is just cheerleading for moral judgements. The structure of debate is used not because moral judgements 'really are' debatable sorts of things, but because donning the form of logical argument makes for more effective cheerleading. Regardless of whether moral debate is grounded in any debatability, it is practical to treat it as such... I suppose, the social phenomena of morality has a structural resemblence to that of a priori truths.
But then, I have a lot of trouble with personal questions of "what ought to be done". This sort of grounded empirical view of morality doesn't do me much good in making moral judgements of my own, in determining my own operation. What do?
My intuition is that people generally have an internal sense of moral feeling. Taking these feelings as propositions for logical extrapolation can give internally contradictory results (see: reductio ad absurdums of utilitarianism) because systems of moral feeling do not actually map onto some coherent propositional logic (at least not without effort). I do feel moral feelings... moldable feelings, self-contradicting feelings... Should I seek to resolve contradiction? There is that word, "should"... Have I not escaped the axiomatizing view? Or should I embrace it, taking axiomatization as an axiomatic value? There's another "should"! Ah! Kill it!
Umm... Nietzsche speaks of this...
(The free will and agency stuff above is also tied up in this. There is a similar instability in thinking about my own agency, I think. But I'm really losing track of myself, now...)
Language: operation vs representation
I have claimed that moral judgements are better understood not as a priori propositions, but social operations, which only take on the structure of debatable propositions within certain contexts. My next move is to suggest that, perhaps, all language works this way?
Thinking of language as representing things in the world works okay for declarative sentences and seemingly nothing else. Rather than understanding language first as representational, a medium for encoding propositions, it seems better to understand it as primarily operational. Even when one is simply stating information (as opposed to asking/demanding), one doesn't do so randomly but with intent to create some effect, be it immediately behavioral or just cognitive. We tell people things to warn them of danger, or to relate, to gain favor, to punish, to argue, to work through disagreements, and so on. We debate in order to persuade, whether about oughts or is's. Representation, then, is an emergent pattern, a structure which forms from operation, within certain contexts. (There is a subset of operations which are isomorphic to representations?)
I don't have very delevoped thoughts on this yet.
Time
Presentism says that what exists is the present. Eternalism (sometimes called "block universe") sees no special real-ness in our present, rather seeing all of time as equally real. When physics became relativistic, presentism was revealed as an anthropomorphic contrivance, it seems. Folk intuitions are so used to conceiving everything as a thing inside of time, that they have great difficulty imagining time itself without erroneously trying to imagine it too inside of time. The common mistaken description of eternalism is that "all of time exists at the same time," which is obviously incoherent. All of time exists, not at the same time, but at different times.
Physics' concept of time as a "4th dimension" has trickled out into the consciousness of laypeople through obscurantist popular physics media which tries to depict everything as whacky and spooky for entertainment. But the math is pretty simple, so I will quickly explain it. The conceptual significance of 4-dimensional "spacetime" is that space and time together have certain symmetries: You can staple any set of parameters together and call it one thing if you like, but spacetime is deeper than that. In space, there are 3 orthogonal directions \((x, y, z)\). But everyone knows that the choice of labeling the directions is arbitrary: one woman's \(x\) can be another woman's \(z\) and so on. In general, we can rotate our axes however we like to get another equally valid set of axes, with no 'correct' choice of coordinates forced onto us- space is rotationally symmetric. Rotations are defined as (continuous) transformations which preserve distance- if I spin around a pencil, its length won't change, as opposed to other kinds of transformations (squashing and stretching). Distance \(r\) can be written in terms of our axes with the pythagorean theorem: \[r^2=x^2+y^2+z^2\] To graduate from space to spacetime, we have to introduce our time axis \(t\) into this equation. Space and time are obviously not entirely interchangeable. But as it turns out, there is fundamentally just this one difference between them: A sign difference. Spacetime's metric is: \[r^2=-t^2+x^2+y^2+z^2\] (Note you can equivalently use \(t^2-x^2-y^2-z^2\). Different physics subfields use different conventions.)
Just as spatial distance is preserved by spatial rotations, this spacetime distance metric is preserved by spacetime rotations, which are called "lorentz transformations" (technically lorentz boosts). Just as space has rotational symmetry, spacetime has lorentz symmetry. Since this isn't euclidean geometry, they look a little different from ordinary rotations, but it's a generalization of the same concept- just as rotations will blur together and interchange the directions of space, lorentz boosts will blur together and interchange space and time, to various degrees, hence "time dilation" and "length contraction". Just as it doesn't matter how you rotate your \((x,y,z)\) axes, it likewise doesn't matter how you "rotate" (lorentz transform) your \((t,x,y,z)\) axes, and different orientations of coordinates will disagree about the spatial distance and duration between the same points.
("Reference frames" are a bad way of thinking about this, in my opinion- it's just different choices of coordinates. When people speak of the "reference frame" of an observer, they are referring to the coordinate system in which the observer is not moving. (Meaning its 4-velocity is orthogonal to the time axis- more on that in a moment.))
The alienness of spacetime to folk intuitions creates all sorts of conceptual difficulties. People often want to talk about "movement through time"; after all, there is movement through space, so if time is just another axis, there should be movement through time, yes? Well, no, because movement is change over time. If one's place in time is changing, what is it changing over? Over time? Really, there is no "movement through time". In spacetime, you shouldn't be thinking of movement at all, you should be thinking of static geometry (which can be framed as movement by picking out the time axis to measure along).
A particle is a point in space which moves along some path over time; equivalently, in spacetime, the path is the particle. Relativistic physics replaces Newton's dynamical models with static geometric descriptions of paths (curves) in spacetime. We throw away 3-dimensional velocity \(v\) in favor of the 4-velocity \(u\), which is just the tangent vector to the path. (Note that tangent vectors are used to capture the direction tangent to the path, so the vector's length is not physically relevant. We normalize their lengths to \(c\) and use natural units (\(c=1\)) so they are dimensionless unit vectors.) 4-Momentum is \(mu\) (with fixed length \(m\)). Since this is a 4-dimensional vector space, momentum has a time component- that's the energy. (If the particle is at rest, its momentum has only a time component, thus \(E=m\), or in nonnatural units, \(E=mc^2\).) Replacing Newtonian "forces", we are interested in the curvature of the particle's path. The curvature is the rotation of the tangent vector, thus "forces" are rotations, which can be written as bivectors (...or antisymmetric tensors or 2-forms, depending on your preference. You can't just use vectors for rotations like in Newtonian rotational mechanics because the cross product only works in 3 dimensions.) Thus the electromagnetic field is a bivector field. (The electric and magnetic fields are its time and space components respectively.)
I'm guessing by this point I've lost everyone who isn't very familiar with physics. This is all much easier to understand if you have spacetime diagrams and equations to look at. But hopefully my explanation is at least somewhat helpful.
Quantum Mechanics (Everettianism)
Even moreso than special relativity, quantum mechanics has made its way into the public consciousness through popular science media which explains things intentionally badly to give the vibe of mystery and paint quantum mechanics as whacky and unintuitive. This is to be expected; what is more strange is how happily even physicists have embraced such silliness. There is the famous Feynman quote:
"I think I can safely say that nobody understands quantum mechanics."
...often misquoted as:
"If you think you understand quantum mechanics, you don't understand quantum mechanics."
This is understandable for 1965, but it's still quoted seriously today. This should be awfully embarrassing for physicists, shouldn't it? Joking about not understanding their own theory?
(Since this is a bold claim for me to make as a nonphysicist, note that I am not introducing anything here as my own radical outsider position. I am siding myself with a position I have seen physicists take.)
Physics is fundamentally interested in how to extrapolate the state of a system forward or backward in time. The "system" is just any collection of stuff we happen to be interested in. The "state" of a system is the way it is at a particular time, which is described by a bunch of numbers (classically, positions and velocities of particles). If you know the state at one time, you can use some laws (Newton's laws or Euler-Lagrange equations or whatever) to derive what the state will be at another time. This is the same in classical and quantum mechanics (and all of physics), but the kinds of states change, and the equations for time evolution change. In quantum mechanics, the state is no longer composed of intuitive things like positions of objects in space; rather, we (roughly speaking) imagine all possible classical states, and assign to each one a number. This set of numbers is the quantum state, the really existing thing, which evolves over time according to the Shrodinger equation. The 'concrete' and intuitive classical framework is revealed to have always only been a statistical limit of this deeper quantum framework.
Example: A single particle. In classical mechanics, the state of the system is the position and velocity of the particle, \(\vec x=(x_1,x_2,x_3)\) and \(\vec v=(v_1,v_2,v_3)\), with some particular values for a particular state. The state changes over time as the particle moves. In quantum mechanics, rather, the state \(\ket\psi\) of the system is a distribution over all possible positions of the particle. (Really, the state is vector in a vector space, where the states with specific position form a basis. It turns out we don't need to explicitly include the velocity/momentum, as it can be derived from the shape of the distribution- it is an alternative basis for the same vector space.) If we call \(\ket{k}\) the state where the particle is at position \(k\), then the any state can be written as a sum of \(\ket k\) states: \[ \ket\psi = \sum_k \psi_k \ket{k} \] ...with the \(\psi_k\) values as the parameters which evolve over time (according to the Shrodinger equation).
Now, if we had two particles, we wouldn't give each its own distribution. The state of a 2-particle system is still a single distribution, now assigning a number to every possible way the two particles can be arranged. If the state where the first is at \(j\) and the second is at \(k\) is called \( \ket{j,k} \), then a general state is a weighted sum of \(\ket{j,k}\)'s.
The difficulty comes with "observation". Quantum mechanics says a system is a distribution over all these 'concrete' states. But when I look at something, I don't see a distribution over all the possible ways it could be. I see it one way, the way that it is. Which state you see can be predicted with the Born rule, which just says the probability of each is the square of its \(\psi\) value. This is not a subjective view, it is a tangible measureable change of state: When you measure, the system really truly changes from some \(\ket\psi=\psi_1\ket 1 + \psi_2\ket 2 +... \) to just a singular \(\ket\psi = \ket 1\) or \(\ket\psi=\ket 2\) or whatever. But how does that work? Why do things suddenly act differently when I'm looking at them? And what even counts as "looking"?
The dominant view among physicists has long been to just ignore this, I think, taking a sort of sloppy dualist or subjectivist or "who cares" stance and then focusing on other things. In practice you can usually get along fine making accurate predictions by just treating observation as a fundamentally separate sort of thing, with its own behavior. But formally, quantum mechanics is not coherent if it can't coherently tell you how measurements are distinguished from other phenomena- this is called the measurement problem. Physics must learn the lesson of 2nd order cybernetics, that the observer is not some special "outside" thing but a system in itself, in interaction with the system being observed, together forming a whole which obeys the same fundamental laws as the individual parts. Quantum mechanics can't get away with using fundamentally contradicting laws for systems of particles than for observers, while also claiming that observers are ultimately systems of particles.
In 1935, Erwin Shrodinger tried to demonstrate how silly this all is via a thought experiment with a cat, a reductio ad absurdum. Say there is a cat in a box (assume the box is idealized so that the inside and outside cannot affect one another). The box has a radioactive element set up so that the quantum state becomes \(\ket\psi= \ket{\text{decayed}} + \ket{\text{didn't decay}} \) (ignoring the coefficients for simplicity). This decay is tied to a poison gas which kills the cat; thus the state becomes \( \ket\psi= \ket{\text{decayed, cat died}} + \ket{\text{didn't decay, cat alive}} \). How ridiculous, Shrodinger says, to think a cat could be both dead and alive, until one opens the box and it is suddenly resolved!
The solution, I think, was basically figured out by Hugh Everett in 1956. The way forward, as it turns out, is not to reject quantum mechanics in favor of some more classical intuitive picture; rather, it is to go further in, to see the quantum explanation through to its natural conclusion. The state is \( \ket\psi= \ket{\text{dead cat}} + \ket{\text{alive cat}} \). If we believe observers are quantum mechanical just as everything else is, and there is no special observation principle, then what should happen when one opens the box to look inside? The answer is: \[ \ket\psi= \ket{\text{dead cat, observer sees dead cat}} + \ket{\text{alive cat, observer sees alive cat}} \] This is the core insight: From an observer's perspective, observations result in apparent collapse of mixed states into pure states, because the observer herself becomes a part of the mixed state. Thus there are, in effect, two observers who see two different things; it is as though they live in different worlds. We don't need a special rule for observation, we just need to take seriously that the observer is herself a part of the system, evolving quantum mechanically.
There is a lot more to say about this with the Born rule and decoherence and pointer states and so on. But I think I've rambled long enough.