Crossposted from the AI Alignment Forum. May contain more technical jargon than usual.

[Metadata: crossposted from I'm fairly likely to not respond to comments promptly. If you're especially interested in chatting, my gmail address is: tsvibtcontact ]

What's a thing, in general? Minds deal with things, so this question comes up in trying to understand minds. Minds think about things, speak of things, manipulate things, care about things, create things, and maybe are made of things.


Examples of things: table, elephant, carbon atom. France, Martin Luther King. Insertion sort. Chess. Unicorn. Learning. Seven. Towel, strand, wing, crystal, finger, space, diffraction. The laws of electromagnetism. The first World War.


What about non-things? It might be hard to list non-things because what we have words for, tend to be things. Redness seems like sort of a thing, but less so. Laws of physics also. Also ghosts. One might say "the ideal gas law is totally a thing" or "ghosts aren't a thing", though I think ghosts are a thing. Events can be things; WWI seems like a thing to me. But a minute ago I picked up my bottle of water and drank from it; that's clearly an event, a real one, but it doesn't feel that much like a thing. The abstract [drinking from a water bottle] feels like a thing though.

(There are some usual critiques of thingness. Yes, there's no sharp dividing line between a wave and a trough, but clearly waves are a thing. Yes, seven is not a physical object you'll ever bump into, but it's clearly a thing. The ideal gas law isn't localized in space or time, but it's a thing. Unicorns aren't real, in that you'll bump into people speaking about them but you won't bump into unicorns themselves and won't be constrained by unicorns in the way you're constrained by seven, but they're things. The world of Ender's Game isn't real, though it's a remote possibility, and it's a thing.)

Features of things

  • Coherence. A thing usually has properties, parts, aspects that are coherent with each other. Some things are big, some things are small; but usually there's nothing like a table that's both big and small.
  • Prediction, homogeneity, constancy, constraint. A thing usually has some predictive meaning; its presence implies constraints on other things or the future, and it implies some constancy in some features.
  • Expression. Related to prediction, a name that names a thing or an idea that's about a thing contributes to expressing thoughts about situations involving the thing. Expressing thoughts about the thing contributes to further behavior such as successfully predicting or manipulating the thing.
  • Cluster. A thing that's an instance of a type of thing, has multiple features mostly shared by most things of that type and not mostly shared by most things not of that type.
  • Exterior, relations. A thing usually can "impinge on other things from the outside". "Reality is that which, when you stop believing in it, doesn't go away."
  • Indicatability. A thing can be indicated. A physical object can literally be pointed at, seven can be pointed at by showing how to count and showing sets of seven like things, and many things can be pointed at by saying the word for that thing.
  • Truth, solidity, fixed import. "True" comes from PIE *deru- ("be firm, solid"), cognate with "trust" and "tree" and possibly "durable". Things have truth to them; a thing makes a promise of agreement between minds on questions about the thing upon further independent investigation. Beyond indicatability, the truth of a thing implies that the thing can be advantageously treated and used in a fixed way.
  • Hollowness, cavernousness. The more comprehensively the history, features, changes, components, causes, internal relations, overarching structures, and explanations of a thing are kept in mind, the less there's a need to think of "the thing itself". One can stand under the cavern more thoroughly, but one never arrives at the thing itself.
  • Interior, promise. A thing usually has something going on with itself, separate from other things; it has internal elements or dynamics. And it promises that it's hiding more depth, there's something further to be understood by looking into the thing more.


How is it that things can be indicated? I can point at a table and ask "where did that tree grow?", and after a moment of confusion and an hour of investigation, you might be able to find an answer; and most people given the same task will probably find the same answer, if they find one. The indicatability of things is strong enough that a straightened finger indicates the table, the table indicates its wooden material and manufacturer, which indicates the lumber company, which indicates a forest, which indicates a location.

Clusters of features are a kind of indication. If the features {wooden, 3 feet tall, has small objects on top, nothing under the top except legs} are perceived together, that indicates the features {flat top, is called "table" by people}. "Indicate" comes from a root meaning "to point out", which involves an agent; that assumes too much, the table doesn't need someone to point. Instead I'll say "reference", in the etymological sense "re-fer", "carry back". If there are clusters in the product space of all features, then specifying the values of some features consistently with a cluster refers you, carries you back, to the rest of the cluster.

Inadequacy of "clustered"

Clusteredness, though, doesn't cover all the thingness of all things.


I think seven is a thing, but it's not a cluster; there's only one seven. Maybe not, though, maybe there are many sevens, which form a cluster?

  • , , , , , , , , ...
  • , a spiral with seven arms
  • the term 7 in Peano Arithmetic, or Robinson Arithmetic, or , or , or some non-classical logic, or some type theory
  • the finite category
  • whatever humans mean when they say "fetch seven apples"
  • the number after six
  • the class of all sets bijectable with { . . . . . . . }
  • 0b111
  • the fourth prime
  • VII
  • the number of Snow White's dwarf friends

Yes, it's not wrong to say that these are many sevens, but (1) the way in which the sevens form a cluster involves features that are heavily dependent on interpretation (e.g. partial translations between logical systems), and (2) the sevens are connected by a lot more than just forming a cluster in featurespace: they play analogous roles in their respective contexts, so that passing between contexts preserves Seven-ness.

Unique things

I think WWI is a thing; it's clearly a member of various clusters like "war" and "20th century human event", but it's unique, and what makes WWI be WWI isn't that it's a member of certain clusters (even if its cluster memberships locate it uniquely). (We can interpret even unique things as clusters: A fabrege egg is a cluster in the space of possible sense-perception-moments, and seven is a cluster in the space of possible thoughts. To me this seems to do some violence to the thing; the thing gives rise to the cluster, but there's more to the thingness than the clusteredness.)

Even if seven and WWI are single, unique things, they have a structure. Seven involves:

  • algorithms (making seven of something, performing addition)
  • features ("is in bijection with [.......]")
  • context (a language, a mind, a logical system, a computer operating system)
  • mental events (picturing seven things, counting to seven, saying or hearing "seven", remembering that there are seven apples)
  • symbols ("7", "seven"), expressions (5+2, 21/3)
  • physical objects (seven of something, neurons thinking about seven, electrons arranged in capacitors in computer memory, sound waves emitted by someone pronouncing "seven", the letters s e v e n written on a page, whatever genes control Graptopsaltria nigrofuscata's allegedly seven-year lifecycle (Lloyd, Dybas))

Inductive nexus of reference

What relates all the stuff in that list? How do these things constitute seven as a thing? It seems like there might be a lot more here, but this essay just wants to describe the situation like this:

The thingness of a thing lies in its being an inductive nexus of reference.


By reference, I mean any re-fer-ence, any bringing-back. X references Y if, when minds think about X, they soon after think about Y. This notion of reference doesn't assume (or exclude) any asymmetry between X and Y: often X references Y and Y references X, as in, the table references supportiveness and supportiveness references tables. There's no separation between stuff that can reference stuff and stuff that can't reference stuff; types and tokens, symbols and meaning, idea and reality, sign and signified, can all reference each other. (This formulation is agnostic about why thinking about Y follows thinking about X. Thinking of X may cause thinking about Y, or they may have a shared cause, or it may be useful to think about Y along with X regardless of the psychic causality.)


By nexus (cognate with "connection", "annex", more distantly "node", "knot"), I mean a bundle, a knot, a region in the graph of reference that's particularly clique-y, particularly highly connected. The shape of the table, the joins between the wood of the leg and the wood of the top, the motions of the carpenter's tools in making the table, all reference each other.


By inductive, I mean that the nexusness is inductive: the highly-connected-ness in the reference graph can correctly be taken to indicate that further investigation would reveal even more highly-connected-ness.

We can notice that dogs share many features with mice and lions that aren't shared by spiders, octopi, and trees: warm, furry, four-limbed, social. Then we can correctly predict that on further investigation, {dogs, mice, lions} will share many further features also not shared by the other living things: red blood cells, a spine, a womb, a spleen.

A non-cluster example: If I hear someone say "seven", I might send blood to the neurons that orchestrate a visualization of seven objects. If I see seven objects and wonder why I'm seeing what I'm seeing, it might be helpful to imagine some process creating seven objects. If I count seven things, and count another set of things, and then announce the total count, I make use of [seven, the thing that can be added]. When I reflect on all this activity, I find that my thoughts can be related to each other, and can be followed "inward, to seven-ness", to think about seven as a real number, as , as iterating processes seven times, and so on. I find connections between all these things. acts transitively on , "because" is a maximal ideal of , "because" you can't fairly give out 7 apples to multiple friends. All this about seven is tightly interconnected, predictably from first noticing some of the connections.


Amusingly (to me), this essay is trying to demonstrate that thingness is a thing. Note that since things have thingness, so that their nexusness is inductive, family resemblances aren't mere.

A Thing is a cavern

So a thing, a nexus of inductive reference, is like a cavern encountered while wandering in a world of caves connected by tunnels. A cave is a small pocket of empty space with a few narrow "in/out-going" passages, on a background of rock. Entering the cavern, at first it seems like just a cave, but stepping in a little further, you find a densely connected maze, many paths connecting the empty spaces; not an empty space on a rock background, but more like a background of empty space contained by walls and broken up by some pillars and stalagmites and stalactites and boulders and rubble.

(Like the cavern, things are strongly (i.e. bidirectionally) connected subgraphs of the graph of references, not just weakly connected. Something that refers to a thing but isn't referred to by the thing, is a mere appearance. That time a few months ago when I took a sip of water from my bottle refers to drinking in general, water in general, hands in general, my hands, and so on, but none of those refer to that particular time I took a drink.)

Talk of "things"

The thingness of things is what we mean when we talk about some stuff being a "thing".

If we say "that's not a thing, that's just some random stuff", we're saying there's nothing to find by investigating that stuff, there's no use in that conjunction of stuff, there's nothing useful or interesting about other stuff that's positioned like that stuff is position, there's no internal structure to that stuff that's relevant to anything other than the separate internal structures of each thing in the assembly called "stuff".

If you ask me "pass me that thing" while pointing, we're relying on the indicatability of the thing, on the reference structure leading from a finger to some stuff, some particular stuff, all of it, and not too much other stuff, all of which I'm led to by the thingness of the thing fairly reliably. You can even ask me "pass me that thing" without pointing, relying on the thingness of the thing to make it jut out ("exist" = "stand out") at me as the thing at the bottom of the basin of attraction of attention set up by our current context. More thingy (at least in a shallow way, and relative to our current context) than the other things, which you would have named or pointed at.

The thingness of [thingness itself] is part of why we talk and think about "things". The activities [dealing with one thing] and [dealing with a totally different thing] have something in common, that is, some Thing in common, namely, the thingness of what is dealt with. So there's transfer of skills between those two activities. And that transfer is usefully inductive: one can learn to [learn, from dealing with one thing, to deal better with any other thing] more effectively than default by investigating the thingness of things, as a thing itself, expecting it to have more referential structure than is currently visible / useable to you, i.e. expecting to be led into more insights about what's in common between things.

(IDK what, if anything, "stuff" is.)


"Thing = inductive nexus of references" tries to characterize approximations to (or rings around, or emanations from, or pathways toward) Kantian noumena, things in themselves. Wikipedia's Kant says: Noumena can't be directly perceived and can't be known, they are completely eternally external to and separated from minds. We can understand the structure of phenomena, which are the appearances of noumena, but we can't access noumena. Noumena must exist because there has to be something that appears to us, an object of investigation, something that we think about.

A noumenon as a nexus of reference is an abstraction over the inductivity of its nexusness: to say "there's a noumenon behind these related phenomena" is to say "so far we've seen some phenomena (appearances) which point to a nexus of reference, but there will always be further (deeper, more, bigger, tighter) nexusness of reference to be found, no matter how many additional related phenomena might appear later on". Noumena say, "what you have is permanently incomplete".

It's maybe like infinity: to say "there are infinitely many natural numbers" is to say "there will always be more natural numbers that we haven't already seen, even if we see more natural numbers later on". Infinity abstracts over the inductivity of the succession of natural numbers. The fact that things-in-themselves / noumena live in a sort of "remote exterior" from our experience or mental grasp, comes from the use of the concept of "thing-in-itself". The concept of "thing-in-itself" is specifically about that which we haven't already grasped, maybe kind of like how infinity is greater than all natural numbers because "infinity" is used specifically to discuss what's beyond any natural numbers already considered.


What would a non-thing be? It would be someth--... some stuff, some event or some happenings, that doesn't have anything more to say, and doesn't separate itself from its surroundings by being itself, is uninteresting, doesn't lead anywhere.

A region on a large wall that's meaninglessly shaped, doesn't encode anything by its position or shape, doesn't surround anything at all special in the wall different from any other part of the wall, would be sort of a non-thing, though it would be hard to produce without making it a thing.

Maybe an example would be a particular instance of perceiving red. Redness in general, and perceiving redness in general, and whatever caused the particular perception of red, would all be things. But the perception of redness itself doesn't seem to offer much; it's at most an appearance of its cause, pointing to the traffic light or the sunset or the anger or the cardinal or the raspberry, and an appearance of perception of red in general, but it's got nothing to it itself, and isn't referred to by those things it refers to.

Maybe combining unrelated sense perceptions also gives non-things; {redness, a slightly flat C# tone, the sense of roughness on your chin} could co-occur but never co-occur again, co-occur for no referential reason, and cause no thoughts beyond what's caused by the individual sensations. Blue is a thing, green is a thing, but grue and bleen not so much; if you investigate bleen, you will be led to blue and green and color in general, and not vice versa, unless you're a philosopher.


I don't like that this notion of thingness relies on reference, which relies on minds, making it seem subjective. At least it can be made maximally intersubjective by saying "minds in general are led from X to Y", which is kind of like being objective.

New Comment
35 comments, sorted by Click to highlight new comments since:

I'm uneasy about this whole thing.  (The phrasing is not a deliberate joke.)  I think "X is a thing" is a neologism, which I encountered on the internet and in Buffy the Vampire Slayer.  "Thing" is a word people use when they haven't figured out a more specific word to use (which is why I used it in my first sentence; I could have said something like "effort to characterize a word", but I might get your intent wrong and don't want to presume, nor want to scrutinize the post carefully enough to choose a term I'd be certain about).

Serious intellectual writings will generally do the work to figure out the right term, and therefore won't say "thing".  The Buffy characters who talk that way are, by doing so, communicating less sophistication or thorough thinking than they would if they used more specific terms, and this fits their characters.

Looking through the examples, I see some very different categories.  Listing a few:

  • physical objects
  • historical events
  • computer algorithms
  • ...

It seems to me that you're trying to use one word to refer to all these different categories, and then do serious reasoning about the word.  This strikes me as a recipe for trouble.

Riffing off your "insertion sort" example, let's consider "Half-Ass", a procedure where you do insertion sort but stop halfway through the list.  If I were telling people that Half-Ass was a sorting algorithm, I think you could say "That's not a thing, there are many input lists that don't output as fully sorted."  But I could code up Half-Ass and run it on my computer.  So we might say, Half-Ass as a sorting algorithm isn't a thing, but Half-Ass as a well-defined, executable procedure is a thing.

It looks like "X, as a Y, is a thing" basically means "X is a member of the Y category".  Then, perhaps, "X is a thing" has an implicit Y, which the audience hopefully correctly guesses.

Using the example of ghosts, "Ghosts aren't a thing" would be correct when the implicit Y is "observable real-world phenomena", but possibly incorrect when Y is "reasonably well-defined concepts" or "phenomena in your favorite fantasy universe".  (I note that some adjacent non-neologism phrases would be "Ghosts aren't real" or "Ghosts aren't a real thing".)

There's even room for debate for something like "Forks are a (real) thing".  Clearly they're mundane real-world objects; but as HPMOR points out, the physical boundaries of a fork are blurry if you look closely enough, it's constantly exchanging atoms with its environment, even its parts have probability wavefunctions instead of actual locations, etc.  So the implicit Y would have to be something like "observable clusters of real-world phenomena"—but no one is going to spell that out except in very rare contexts.

One thing I could say is, anytime you get really deep into examining a statement like "X is a thing", you should note that the word "thing" is a signal that the speaker put far less effort into being precise with their word choice than you are putting into analyzing it.  Or, if you're considering "When would I say that X is a thing?", you're thinking about it much more carefully than you would in most actual scenarios where "thing" was the best word you could come up with.

It is somewhat interesting to think about how precisely people word their statements, under what circumstances; and interesting to think about the advantages of being able to state imprecise thoughts early in an investigation, versus the advantages of forcing yourself to make your ideas precise (cf. Orwell), and maybe how to gain from both; and there are questions about categories and fuzzy boundaries and stuff (which I feel were adequately explored in e.g. "The Categories Were Made For Man").  But you should be clear about just what it is you're studying, and it's probably better to disentangle the issues of (justifiably) lazy communication from the other issues, and study them separately.

But you should be clear about just what it is you're studying

The post discusses the thingness of things. Seven, for example, seems like a thing--an entity, an object. I naturally mentally relate to seven in many of the same ways that I naturally mentally relate to a table. So the question is, what if anything is in common between how we relate to each of those entities that seem like things?

Half-Ass is a reasonable example of a somewhat non-thing, according to the hypothesis in the post. It refers one fairly strongly to "half" and to "ass" "insertion sort", but "insertion sort" barely refers one to Half-Ass, and likewise "half".

Hmmm... perhaps the concept you're going for is "a thing that my brain thinks is worth remembering (and/or categorizing, naming, or otherwise having a handle for)"?  Which, of course, would be highly subjective and context-specific.

Suppose you were teaching CS students, and it was for some reason very common for your students to make that specific error of replacing "i < length(A)" with "i < length(A)/2", because ... maybe you're working with a high-level language, but you're implementing this part in assembly for speed, and the language runtime actually represents the integer N as 2*N because of pointer tagging and it's common for students to forget that part; or maybe this is a compiler bug, triggered in rare but known circumstances (e.g. when the compiler decides to put the length into a certain register which is treated specially).  Then you find it useful to know how Half-Ass behaves, so you can better test and diagnose your students' programs, or create a test case to detect systems with the buggy version of that compiler (perhaps for white-hat or black-hat security purposes).

In those scenarios, does Half-Ass seem like more of a "thing"?  And if those scenarios were real, but also there were plenty of "civilian" programmers who'd never used that language or that buggy compiler and probably never will, would it be valid for those programmers to say "No, I don't think that qualifies as a thing"?

It [Half-Ass] refers one fairly strongly to "half" and to "ass" "insertion sort", but "insertion sort" barely refers one to Half-Ass, and likewise "half".

Hmm, you seem to be attaching significance to the words I used in the name.  I would have thought that "whether something qualifies as a thing" was mostly independent of whatever words people had come up with when trying to name it.  (The most descriptive name would be "insertion half-sort", incidentally.)

In those scenarios, does Half-Ass seem like more of a "thing"?

IDK, but I like the question.

I'd say that what does seem like a thing is [insertion f-sort] where the fraction f is a parameter. Then [insertion 1/2-sort] is like [this particular instance of me picking up my water bottle and taking a drink], and [insertion f-sort] is like [me picking up my water bottle and taking a drink, in general].

Unless there's something interesting about [insertion 1/2-sort] in particular, like for example if there's some phase transition at 1/2 or something. Then I'd expect that it's more of a thing (for example, that there'd be further interesting special properties of [insertion 1/2-sort] that we haven't discovered yet, or that there'd be analogies to other phase transitions).

In the compiler example, the compiler representing N as 2N is more of a thing.

[insertion 1/2-sort] is then a somewhat meaningless coincidence of other meaningful things. You have your symptomatic cluster of phenomena: the CS student complains that a check fails, but when they print out the list, it seems sorted (though only the first part of the list actually printed, before the summarizing ellipsis....), the algorithm is clean and obviously right, etc. There's some thingness I guess, in that there's an insight to be had. But it's a cavern which, when entered, turns out to be pretty clearly two other caverns connected by a tunnel. (I admit that the dynamical story here is unrefined, and I'd be interested in a better picture.)

(sorry for the unnecessary length. this is a thinking-out-loud sorta comment.)

pretty sure things are always physical shapes, even when they insist really hard that they aren't; I'd also say that all physical stuff is part of some thing, but that the natural thingness abstraction scene graph of nesting of things may not be very deep for some stuff, such as the inside of the sun, which probably has relatively few discretely referenceable, thing-grade patterns, just a bunch of stuff happening.

anyway, let's see how "things are when stuff's shape is coherent" holds up on your test cases.

  • table: a set of shapes of physical object; an object is a set of shapes of energy in mass fields.
  • elephant: also a set of shapes of physical object.
  • carbon atom: also a set of shapes of physical configuration.
  • France: a different kind of selector, but still a set of shapes of physical configuration.
  • Martin Luther King: a person; people are shapes of physical objects; a person who is gone, so in other words, the physical object has diffused away and what is meant now is that person's impact on the world, such as writing, speeches, etc; those writing and speeches are also parts of physical objects, but they're incoherent in spacetime, in contrast to the examples so far. this is only because MLK is gone, and any other physical shape that is gone would have the same pattern.
  • Insertion sort: phew okay not sure about this one chief. I'm not sure insertion sort is often a whole thing unto itself. it's almost always part of something else. but I guess you can say that the physical shape of insertion sort is shared by all instances of insertion sort. insertion sort is therefore a pattern in the surface of other matter.
  • Chess: i guess it's a shared shape of physical configuration, sure. its precise definition would include matcher-generators for shared pieces and unshared team identifying parts, but ultimately all instances of it take a certain relationship in their physical shapes. it's still a kind of physical object, but it's abstract over type of parts instantiating it, just like insertion sort.
  • Unicorn: so far so good, unicorn is also a description of a shape of physical object. the way the set "unicorn" is defined accepts images as description of a unicorn's shape. There are no species of genetically descended thing that quite match the attributes of a unicorn's shape, but things that qualify as a unicorn and images that qualify as images of a unicorn have been made by human artistic endeavor, and both of those are physical things. it would be possible to shape genetics so that a unicorn instantiated with genetically descended parts could become real. (rhinos might be upset about this.)
  • Learning: okay so, learning is only a physical object if you accept, as described above, that algorithms are themselves patterns in the shapes of physical objects. learning is a set of changes of an object's shape in order to make some inner part of the physical object share patterns with some external environment, potentially in ways steered to make the object's shape have effects outside itself.
  • Seven: a basic shape of physical matter that can be made of many kinds of part; numbers are a very very general description of shapes of matter. Seven has to be of a type of part, such as successor function nestings on a page or screen or ram chip, or of cells. notably, this means that "seven", "7", etc are not seven the thing, they are references to seven. seven is satisfied by [. . . . . . .].
  • Towel: another complex grammar of shapes with default settings.
  • strand: a complex grammar of shapes but significantly less complex than a towel; notably, almost always a subshape of a towel. a component thing.
  • sidenote: towels also contain a lot of sevens (sevens of atoms, or sevens of molecules, or sevens of strands...)
  • wing: yep, physical object, covered by examples so far.
  • crystal: much more precise set of physical objects.
  • finger: see wing.
  • space: physical shape that is "almost no energy in this window"? or alternately, any physical shape within a given window or something.
  • diffraction: oh man uh I guess this is a shape?
  • The laws of electromagnetism: this is not a physical object no matter how hard I stretch it, oh man. I don't think this one belongs in this list, actually. or, well, I guess the laws of electromagnetism are still a set of weighted possibilities of physical shapes, so maybe the ontology still works?
  • The first World War: shares attributes with france (large aggregate network of shapes comprise it as a thing), MLK (gone now, had lasting impacts)

coherent, potentially-reinstantiable shapes of physical objects with heavy self-connectivity. thingness is probably some form of clustering of repeatable patterns of components, such as how a towel is made of many possible grammars of strands and sequences of atoms.

seems like it would help to separate instances of things, references to things, and detailed description of the things sufficient to make a new instance, but which does not itself qualify as an instance of the thing. for example, "seven" is not a description of how to make a new instance of seven. ("four" is the only word lucky enough to describe itself in english; on the surface of your screen is the thing four in every place the word four shows up.) I'd also say 7 is not the thing, it is a reference. the laws of electromagnetism are a concrete physical shape which is instantiated everywhere, but the way we write down the variations allowed by that shape is a symbolic manipulation of patterns allowed by the math that describes so reliably what we see upon experiment. (and hence why we suspect that these near-perfect grooves in reality may add up to fundamental sameness such as living on a quantum turing machine or whatever.)

(I want an anti-repetition tool omg I repeat myself as much as any llm ugh)

I don't think we know to what extent abstract objects need to share isomorphic neural structures that represent them in different individuals.

To present an extreme example, imagine a human and a hypothetical alien, the human and the alien both share an emotion that we can categorise behaviourally as "love", but their neural structures are unrelated. In both cases we can use the same concept of "love" to make predictions and inductive reasoning. The concept of "love" applies to both.

Should we think they share the same relevant neural structures to share the concept of "love"? I would not expect that, entirely different neural structures could conceivably produce 'things' that are clearly analogous, we could say that here there's no one 'love', but two different emotions, but we can say the same thing about many other 'things' that we consider one thing or their own, like different designs of table.

This might be the case for mental representations in humans. To what extent should we think that two humans represent things by having the same neural structures? I don't know.

but there must be a shared property of both physical shapes in order to describe that property's dynamics and call it love, yeah? as long as we limit ourselves to describing physical data using our abstractions, the impact of being abstract is that the shape of the concept is built out of a grammar that generates and accepts more structural variations - the grammar defining the concept "love" presumably involves component features such as "agentic caring", which is a feature of action patterns within and between brains, those actions themselves a spacetime shape of energy transfer, or of "enjoyment", a feature I would guess is defined by the satisfaction of agentic caring, which is to say that the networks of pattern in whatever brain you happen to have implementing agentic caring about another brain are in a state of satisfaction within the range of target shapes they'd like the cared-about object to take, this satisfaction in turn defined by lack of intervening action generated by the other brain's state according to the evaluating brain.

obviously love has a lot of parts and we're always hesitant to ensure we don't miss parts, as love is a fuzzy thing and we're never totally sure we've labeled it right.

but to the degree it's the same, it has to have some sort of structural similarity. abstraction exists In terms of what patterns we can call the same due to shared features of their grammar.

Do they have to share a structure though? I was trying to present a possible exception. Part of my point was that evolutionary processes might be able to create two different processes that are behaviorally equivalent. It could be something analogous to two machines giving the same output through entirely different inner processes.

'Love' is ultimately just a high-level description of behavior in living systems, people are even arguing how much our folk mental concepts are accurate descriptions of brain states.

There's also the fact that we can seemingly abstract as much as we want the precision of structures to define 'things', specially with artifacts, like tables (because of the metaphysical concept of 'purpose'). If we unrestrict composition, that any two instances of what we call the same thing share a common descriptive basic structure becomes trivial. I believe this is part of the reason numbers are so 'useful', they are such generic/basic/undefined constructs that is difficult to imagine an universe in which there's no stuff that we can abstract into things to count.

If we don't want the idea of shared structure to be trivial, we need to restrict the idea of 'thing' to something that is 'natural' instead of 'social'. For an emotion to be 'natural', it means that it's an accurate and consistent description of a brain state separate from other so-called 'emotions' with somewhat defined boundaries. I'm assuming a computational theory of mind.

Then, if we have 'human love 'and 'alien love', it might just be that the neural structures that identify 'alien love' are not anymore similar to 'human love' than to any other human emotions. Then the thing that they would share to be called "love" would be their behavioral effects, their effects on the peripherals of the brain-body system and their subsequent behavior. This might be possible through the process of convergent evolution; if it's not possible, that means there has to be a shared computational structure for the same behavioral program called that we identify as 'love'.

In a theory of embodied mind, the peripherals would be part of the mental process of 'love', so they would probably need to share a common structure by necessity.

You might be interested in Quine's collection of essays Theories and Things, where he defends some version of "things as space-time regions". I'm pretty skeptical of your version though; or at least, I'm interested in why 7 seems like an object.

chess is a grammar for physical systems, which I started trying to write out, before realizing I don't know the rules well enough. but anyway it defines a network of position representations connected in a grid pattern with state transitions constraints; the chess grammar can be implemented by many physical substrates, eg a physical board with a grid on it, or a list of randomly shuffled board location names on a whiteboard - changing the projected geometry doesn't change the game unless the movement metric changes, so I'd still classify the aggregate physical system as chess. It could be quite valid to argue that the variety of possible instances of that grammar includes many physical systems which are not single objects, due to binding the word "object" only to physical systems that have connected internal molecular bonds or such things.

I'll have to check out the reference.

Also, I don't intuitive understand scientific laws as instantiated on physical objects by themselves. This might just be a double standard, but I understand scientific laws as epistemic states or descriptions of phenomena, so they are contained as information in brains and machines.

You don't need to tell an electron how to electron but it happens on its own. An electron doesn't proton, there are personalities/dynamics that it eschews and this is happening restrictive.

When you try to write claims in prose of english there is nothing restricting you from writing an inaccurate description in contrast to an accurate one.

For a complicated thing the only human endeavor that knows about it is a scientific one. There are not a lot of doorsellers that need to refer to binary stars in their job. "doorsales law" could be that getting your foot in the dooor before it closes increases conditions under which a sale happens. "scientific" in that sense is the dealing with the weird just for the sake of eruditeness (kind of a synonym for "paranormal").

"Epistemic descriptions" is the better meaning, imo.

perhaps laws are the writing. but they describe a reliable pattern in the shape of the fields; the law is the math expression that traces the measured shape of the field out on a piece of paper or a screen, perhaps. to look up visual examples of the shapes I'd look up electromagnetics measurement experiment demonstrations on youtube or so. there's all sorts of good rf visual demonstrations.

Is it fetch, yet?

I very much appreciate trying to figure out what things are. I think, though, you've added more complication than needed. However, my take depends on a particular view on philosophy.

So, first I think Kant is wrong about noumena. They don't exist. There are no things in themselves, there are only phenomena: things that exist because we reify them into existence to fit some concern we have. Things are reified out of sensory experience of the world (though note that "sensory" is redundant here), and the world is the unified non-thing that we can only reify by virtue of it being the whole of existence and it is defined against the null set of hypothetical non-existence.

So given this stance, things are then just us experiencing the world and putting bits of it into little boxes by making claims that a things are this and not that. The non-thing is the nothing of the whole, unified, undivided world.

One very interesting consequence of this view is that things exist only in the map, not in the territory, because things only exist by virtue of some part of the world experiencing itself and creating a little pocket of self-referential information.

I explore this idea in more detail here by considering the special case of causation:

there are only phenomena

Do I only exist because you "reify" me?

This is confusing two different notions of exist. There is existence as part of the wholeness of the world that is as yet undifferentiated and there is your existence in the minds of people. "You" exist lots of places in many minds, and also "you" don't have a clearly defined existence separate and independent from the rest of the world.

I realize this is unintuitive to many folks. The thing you have to notice is that the world has an existence independent of ontology and ontology-less existence can't be fathomed in terms of ontology.

Are you saying my existence is "undifferentiated" from "the wholeness of the world" so long as no one else is observing me or thinking of me?

Yes, though note you can observe yourself.

How can self-observation be the cause of my existence as a differentiated being? Don't I have to already exist as a differentiated being, in order to be doing that? 

Oh, I thought I already explained that. There's at least two different ways "exist" can be meant here, and I think we're talking past each other.

For some thing to exist that implies it must exist ontologically, i.e. in the map. Otherwise it is not yet a thing. So I'm saying there's a difference between what we might call existence and being. You exist, in the sense of being an ontological thing, only by virtue of reification, but you are by virtue of the whole world being.

I have a theory that belief in a good God is the main delusion of western religion, and belief in a fundamentally undifferentiated reality is the main delusion of eastern religion. 

I see no way around the conclusion that differences are real. Experience is part of reality, and experience contains difference. Also, my experience is objectively distinct from yours - I don't know what you had for breakfast today (or indeed if you had any); that act was part of your experience, and not part of mine. 

We can divide up the world in different ways, but the undivided world is already objectively differentiated. 

Sure, differences are as real as the minds making them are. Once you have minds those minds start perceiving differentiation since they need to extract information from the environment to function. So I guess I'm saying I don't see what your objection is in this last comment as you've not posited anything that seems to claim something that actually disagrees with my point as far as I can tell. I think it's a bit weird to call the differentiation you're referring to "objective", but you explained what you mean.

Once you have minds those minds start perceiving differentiation since they need to extract information from the environment to function.

How can there be information for minds to extract, unless the environment already has some kind of structure?

Why does there need to be structure? We can just have a non-uniform distribution of energy around the universe in order for there to be information to extract. I guess you could call this "structure" but that seems like a stretch to me.

I don't know if I can convince you. You seem pretty convinced that there are natural abstractions or something like them. I'm pretty suspicious that there are natural abstractions and instead think there are useful abstractions but they are all contingent on how the minds creating those abstractions are organized and that no abstractions meaningfully exist independent of the minds that create them. Perhaps the structure of our universe limits how minds work in ways that de facto means we all create ontology within certain constraints, but I don't think we know enough to prove this.

By my view, any sense in which abstractions seem natural is a kind of typical mind fallacy.

a non-uniform distribution of energy around the universe

So in the end you are willing to hypothesize that reality has gradients of difference, even prior to the activity of minds? That was my biggest stumbling block, everything else is a detail. 


So, first I think Kant is wrong about noumena. They don’t exist. There are no things in themselves, there are only phenomena

Kant says that noumena don't exist. Kant also says that the thing in itself does exist. From these statements, we can infer that (despite what many people think) "thing in itself" is not a synonym for "noumenon".

A noumenon is something external to the mind that is knowable in a purely intellectual, non sensory, way. (Eg. a Platonic form).

A thing itself is the assumed source or cause of a phenomenon (or intuition on Kant's vocabulary). But there is nothing in the concept of the thing in itself that requires it be knowable noumenally in addition to its phenomenal effects.

The famous statement that the thing in itself is not knowable is another way if saying it is not a noumenon...if it were a noumenon, it would be knowable intellectually. Phenomenonal appearances also do not reveal the thing in itself , because they are only ever in relation to an observing subject ... so not "in itself". Nonetheless, entirely general considerations suggest that appearances must be appearances of something, ie. have external causes


Things are reified out of sensory experience of the world (though note that "sensory" is redundant here), and the world is the unified non-thing

Okay, but the tabley-looking stuff out there seems to conform more parsimoniously to a theory that posits an external table. I assume we agree on that, and then the question is, what's happening when we so posit?

Yep, so I think this gets into a different question of epistemology not directly related to things but rather about what we care about, since positing a theory that what looks to me like a table implies something table shaped about the universe requires caring about parsimony.

(Aside: It's kind of related because to talk about caring about things we need reifications that enable us to point to what we care about, but I think that's just an artifact of using words—care is patterns of behavior and preference we can reify call "parsimonious" or something else, but exist prior to being named.)

If we care about something other than parsimony, we may not agree that the universe is filled with tables. Maybe we slice it up quite differently and tables exist orthogonal to our ontology.


I'm asking what reification is, period, and what it has to do with what's in reality (the thing that bites you regardless of what you think).

This seems straightforward to me: reification is a process by which our brain picks out patterns/features and encodes them so we can recognize them again and make sense of the world given our limited hardware. We can then think in terms of those patterns and gloss over the details because the details often aren't relevant for various things.

The reason we reify things one way versus another depends on what we care about, i.e. our purposes.

I didn't link it in my original reply by work on natural abstractions is also related. My take is that if natural abstractions exist they don't actually rehabilitate noumena but they do explain why it intuitively feels like there are noumena. However abstractions are still phenomena (except insofar as all phenomena are of course embedded in the world) even if they are picking up on what I might metaphorically describe as the natural contours of the territory.


How do they explain why it feels like there are noumena? (Also by "feels like" I'd want to include empirical observations of nexusness.)

To me this seems obvious: noumena feel real to most people because they're captured by their ontology. It takes a lot of work for a human mind to learn not to jump straight from sensation to reification, and even with training there's only so much a person can do because the mind has lots of low-level reification "built in" that happens prior to conscious awareness. Cf. noticing


Are you asking what a thing is, as opposed to a non thing?

"Thing" might be the broadest of all categories...or it might contrast with traditional categories, like, "event", "process", "state", "property", "relation", "type", "law" etc.

Speaking of which...ontology is a thing , in the Buffyesque sense...there are thousands of years of prior art available. So there is no particular need to reinvent the subject from scratch.

Wikipedia’s Kant says: Noumena can’t be directly perceived and can’t be known, they are completely eternally external to and separated from minds.

Errmmm...I think that would be a paraphrase.