## What is computation? Does computation = causation

• 17.4k
A bundle of sticks that looks like this: VIII with no one to observe it is a bundle of sticks. It can't ever be more than that without some mind observing it and attaching additional signifiers. However, when the bundle of sticks is observed by someone who knows Roman Numerals, it's a bundle of sticks AND it picks up a new attribute courtesy of the mind observing it: it's a bundle of sticks and the roman numeral for 8.

Know exactly what you mean. I had a marathon thread here in the past about just this kind of thing. The broader situation is, modernity divides the Universe into subjective and objective. Then it says that the objective domain is entirely devoid of meaning, because meaning resides in the subject. Then it asks, why is it meaningless?
• 891
Here is a demonstration
Another fun version I have thought about before:

A very simple algorithm outputs every possible combination of RGB values in a 1024x1024 pixel image. The program, which can be written in an afternoon by a competent programmer, produces 256^3^1024^2 images. Pretty pedestrian, as far as big numbers go. But this program's output will include:

• Every painting ever painted
• Every painting, with the subjects replaced by anthropomorphic versions of every animal, in every permutation
• Every one of your pages, rendered in every font
• The above, where every character is rendered in every permutation of fonts
• The above, but where every character is also rendered in permutations of every discernable shade
• The above, but where every permutation is rotated from 0 to 360 degrees, to the limits of discernibility
• Every visual experience ever had by every human, awake, asleep, drugged
• And every other animal
• The above, but where the scene is rotated in every discernable degree around all 3 axes
• The above, but where overlaid on top is each and every one of the page images above, with every discernable permutation of transparency, scale, offset
• Every distinct vantage point of every location on earth, at every scale, and every time
• ...and every other planet, and every other visually distinct location in the universe
• ...then overlay the pages
• ...overlaid on top of that, variations where every forum member here is mooning the "camera", at every discernable location, scale, posture, age, skin tone...

...And this only scratches the surface of the surface of the surface of all the discernable images
.,..And yet, the vast, utterly overwhelming majority look like colored dots.

If algorithms are just names, a relatively bare bones symbol shuffling algorithm is almost godlike in it's ability to name almost everything.

It just "names" (or, as we prefer, is another form of) every possible book, which is quite a different thing from any particular book. Selecting a desired book out of this heap is another computational problem, which the algorithms definitely do not solve.
• 1.3k
Exactly
• 768

So something changes in the computer when it is observed or is computation just in the mind of the observer? If the latter, why is it not the same for all universals, e.g., six rocks are not "six" until observed, a triangle isn't a triangle until it is observed, etc? I.e., nominalism.

I don't think I agree. It seems difficult to have information be mind independent but not computation. I won't comment on the status of such things in theoretical "mindless universes," but in the real universe meaning, at least at the level of reference to something external to the system, absolutely seems to exist sans observers, e.g. ribosomes are presumably not conscious but can read code that refers to something other than itself, and they in turn follow the algorithm laid out in the code to manufacturer a protein.

DNA computers organized to produce solutions to Hamiltonian path problems don't behave physically different from DNA in cells at a basic level, so it's hard to see what the difference would be.
• 1.3k
So something changes in the computer when it is observed or is computation just in the mind of the observer?

No, looking at a computer doesn't cause any changes to the computer. However, observing the output of the computational process and attaching significance to it allows us to attach the word "computation" to the whole process. I don't think observation is sufficient to establish computation. I don't think a baby watching a computer counts. It has to be something that understands the output of the computation, something that can attach meaning to the output.
• 768

And is this the case for all universals? I can't say I find that to be an attractive position.

For one, look at the Chinese Room thought experiment. There it certainly seems like one can have computation without understanding. Or, in the example of the China Brain, you would have conscious entities carrying out computations they weren't aware of and couldn't ascribe meaning to.

In any event, I don't think this solves the pancomputationalism problem even if we accept your premise. The pancomputationalism problem/hypothesis arose because conscious philosophers and physicists observed computation everywhere. So the problem is still defining a non-arbitrary definition of computation, even if we bracket computation to just observed phenomena.
• 1.3k
I don't know much about universals. My thoughts on this are very preliminary. I would like to see Wayfarer take this on. I think you've done an excellent job in this thread.
• 768

Isn't this guilty of the same division?

Humans are part of nature. Human minds presumably have natural causes and thoughts/subjective meaning are part of this natural world. The stick representing the number seven is a fact of nature, something empirically observable and testable.

That the sticks don't signal "7" to someone ignorant of Roman numerals or a dog is simply due to the relational nature of information. Having mind "create" new attributes seems to me to be falling into the same sort of (artificial) dualism.

You have the same problem within subjective experience when someone mistakes a fire alarm for a carbon monoxide detector or a burglar alarm. The information is substrate independent, but not arbitrary. If I mistake my fire alarm for a burglar alarm that does not turn the one into the other, regardless of the meaning I take from the signal. I can discern this if I trace the origins of the source.
• 891
My guess is that it comes down to the ability to discern between small differences. This is also what instruments do for humans and computers, allow for greater discernablity.

Imagine an oscilloscope attached to an ethernet cable. Properly tuned, the image it displays will be sensitive to the electrical activity of the cable. Now record the oscilloscope with a video camera, and you have a system which is sensitive to minute changes of the cable over time.

But, recording and displaying this video is all it does. As sensitive as it is, the behavior of the system is still causally driven by the physical activity of the cable. You can understand the visual behavior in physical terms, which is why the oscilloscope is useful.

Contrast that with what happens when you plug the cable into the computer. The signal might be interpreted as an image, or a sound. Or logical instructions which when executed implement a set of abstract rules, such as how to play chess.

This is what I mean when I say that computers are not bound by causal reality. Unlike the oscilloscope, there are no laws of physics that correspond with the rules that it implements. There is no physical system that maps to the rules of chess, it is an abstraction realized by the computer. This I think is the key point that distinguishes computation from causality.
• 17.4k
It seems difficult to have information be mind independent but not computation. I won't comment on the status of such things in theoretical "mindless universes," but in the real universe meaning, at least at the level of reference to something external to the system, absolutely seems to exist sans observers, e.g. ribosomes are presumably not conscious but can read code that refers to something other than itself, and they in turn follow the algorithm laid out in the code to manufacturer a protein.

Interesting that the only place outside human activities and animal communications that something like transmission of information occurs is in living organisms and DNA, isn't it?

Humans are part of nature. Human minds presumably have natural causes and thoughts/subjective meaning are part of this natural world.

The mind is not something observable in nature. We can observe that other creatures are conscious and presume that they too have minds, but the mind is never a direct object of perception.

As for the interpretation of numbers and so on, humans inhabit a 'meaning world'. It doesn't comprise only objects, but also consists of a continuous process of interpretation, whereby we assign meaning to everything we encounter. Within that matrix, what is objective and what is subjective arise together - we don't see the world as if from no viewpoint, although we think it's easy to do so. But even the imagined panorama of an empty universe is organised around a point of view, without which there would be neither scale nor perspective.
• 768

First, isn't it the case that digital computers obey all the physical laws we know about? Hence why knowledge of physical laws has been essential to creating them in the first place.

The oscilloscope doesn't discern the same differences. For example, it is going to present instructions like "open this program and interpret the following in terms of this software," as simply a line, and the instructions as variations in that line. It is unable to discerned how prior signals can change the interpretation of later signals or how later signals can modify the context of earlier ones. Essentially, it lacks an ability to discern differences in signals over time, because it lacks memory.

Obviously part of what makes computers so useful is their ability to take new information/instructions and combine them with information already encoded in the computer itself. So, discernablity of inputs isn't the only important thing, there is also the ability to take on more discernibly different states that, importantly, will correspond to changes in their macrostate.

Computers are very low entropy, which means in terms of a Boltzmann distribution and how their micro constituents are organized, they can take on much fewer distinct microstates states that align with their current macrostate than say a rock or a volume of gas. But what is important is that these state differences are such that the systems they interact with can discern between the state changes, and in turn that these state changes can be transformed into discernible macrostate changes.

Letters or video appearing on a monitor, a human doing a dance or picking up a guitar, these are all discernible macro changes resulting from micro changes. Throw either of these amazing systems into a magical blender that mixes up the constituent molocules and you're extremely unlikely to get any microstates that produce macro changes by chance.

Of course, if you smack a computer monitor and a rock together, the result is identical regardless of what the monitor was displaying. But for us, the output on the screen matters quite a bit. I think this is what you and Wayfarer are getting at. The problem there is that there are plenty of other differences in the physical world we can think of that are completely indiscernible for us, but which radically alter how some other, presumably non-concious system responds to them. This is just the relational nature of information.

Interesting that the only place outside human activities and animal communications that something like transmission of information occurs is in living organisms and DNA, isn't it?

Is this the case? Doesn't water eroding topsoil generate information about its passage in the form of riverbeds? This seems to be why we can comment on the age of the Grand Canyon and the history of its formation, precisely because water encodes information on sandstone. Likewise, we can discover things about the atmospheres of distant planets because changes in light from far off stars due to the interactions between the light and the planets' atmospheres encodes information about these planets.

The Vortex optic/fire control system for the US Army's new 6.8mm rifle can instantly zero itself onto any target. It does this by symbolically representing inputs from a built in range finder and atmospheric sensor, which are then analyzed by the ballistics computer. SHARP, a full fire control solution goes a step further, by recognizing targets for the user. It is able to process symbols that represent something else (the target) well enough that a user only has to hold down the trigger continuously to place accurate fire. The weapon will only discharge on a calculated hit (reducing recoil and decreased volume of fire concerns that come with using a full power cartridge).

It will accomplish this symbolic representation even if mounted to a drone, with no immediate conscious observer.

but the mind is never a direct object of perception.

Aren't our own minds the objects of direct perception? Arguably this is the only thing we observe directly, depending on how you define direct. Light, apples, cars, these are all filtered through the mind, Kant's old trancendental and all.

I'm curious on this line of inquiry though, do you think artificial intelligence could generate such meaning? Do dogs experience it?

It seems to me that this risks conflating the concept of information, which seems to be widely applicable to the natural world, with the presence of first person experience, which is on the one hand everywhere (all objects are subsumed in it) but also generally presumed to only be connected to a small fraction of all the external objects in intersubjective reality. I don't think the former is necessitated by the immediate presence of the latter, although perhaps the existence of information does require the potential of experience.

I say this because I think it's likely the case that light carried information about far off planets to the Earth even before the Earth had life one it. If it didn't, I don't know why we shouldn't just take the extra step of saying the Earth didn't properly exist until life did.
• 768
Actually, after all that, I think I might agree with the basic idea that in a way, information, at least in terms of say, what we get from looking at a written page, doesn't exist when it is not observed. Discernible difference defines information at any one place and time. Information transfer, computation, only occurs locally.

So, I would agree that there is an important sense in which differences/meanings that are only discernible for human beings do not exist when no human being observers them. That is, their existence or non-existence is identical for describing reality for some given period P. But, in an important way the information must exist during P, in that its potential is always there.
• 17.4k
Is this the case? Doesn't water eroding topsoil generate information about its passage in the form of riverbeds?

But that's informative to us. The difference with the information encoded in DNA is that it is morphogenetic, i.e. it causes things to happen, it transmits and stores information. THat is why some (although not all) biologists recognise an ontological distinction between life and non-life - living things are different in kind, not just in degree, to the elements of the periodic table.

Aren't our own minds the objects of direct perception? Arguably this is the only thing we observe directly, depending on how you define direct. Light, apples, cars, these are all filtered through the mind, Kant's old trancendental and all.

Very tricky distinctions, but I say that it's not. The mind is primarily the subject of experience, that which objects are perceptible to. We can't stand outside of the mind and make it an object in the same sense we can objectify perceptibles. We can obviously talk about our state of mind and mental events, but the question of what the mind is, that has these experiences, is a deep one. There's a theme in current phenomenology about this idea, along the lines of 'the mind knows but is not known', as it is always the subject or recipient, never amongst the objects of perception. It's a question with an ancient heritage, and of course, you're right in mentioning Kant.
• 521
We can observe that other creatures are conscious and presume that they too have minds, but the mind is never a direct object of perception.

I tend to question what could be meant with that. We say a reflection on thought is one of our selves but do not overcome the distinction of the observing and the observed. The "observed" thought - the words in "mind" - is "there", in "space". Absolute Idealism hinted that this perception is already adequate. There is no arbitrary determination of things done "by the mind". The things themselves have already imposed their negativity, ie their restrictions on what they can be, on the subject which is a part of totality:
A try of reduction leads to the thought of two "me": It is the pure observance of "I am"-me that has the quality of "I think"-me. We tend to think the subject as the active - which it is (in common sense) only as long as it misunderstands itself as it's object. A camera being moved has the impression to move; the pure observance of "a tripod wanting to walk" bears the impression of a want to walk. How fitting!
• 891

It sounds like you guys are conflating information and interpretation. If these were the same information could not be interpreted in multiple ways. Only interpretation cannot occur without an observer, and this can include machines as well as minds.
• 17.4k
Is there such a thing as uninterpreted information? Put another way, how does anything constitute information until its been interpreted? I mean, the genetic information transmitted by DNA is interpreted by ribosomes. But in the non-organic realm, what sense does it make to speak of information at all? Sure, we can ascertain vast amounts of data about the Universe, which then constitutes information, but does the Universe itself constitute 'information' in any meaningful sense?
• 891
. But in the non-organic realm, what sense does it make to speak of information at all?

Computers certainly operate on information.

Does a library at night have any information? Do all the books have information, or only the ones currently being read?

I don't know how you define information. If it is state, there is certainly state without interpretation.
• 17.4k
Computers certainly operate on information.

Computers and libraries are human inventions. Whatever order they have originated from that.
• 768

Interpretation doesn't need a conscious observer though. Plenty of industrial systems are set up in such a way that the same signal is meant to represent different things in different contexts. This is true in software too.

If it is state, there is certainly state without interpretation.

Exactly. Books in a library have information because of their states. And notably, interpretation is also a question of states. If I tell one person "if you see me raise my hand it means go start the car," and another person "if you see me raise my hand I want you to grab my bag," my act has two different meanings because it is being computed in the context of previously exchanged information that resulted in state changes in my interlocutors.
• 768

I would argue that information exists "in the wild," as discernible differences. If information only existed when observed we would have to posit that observation somehow changes state differences into information, a change in the object, or that all information only exists in the minds of observers. So a riverbed wouldn't store information of the passage of water, but then its physical state, which seems identical to the total information that can be taken from it, is somehow different?

I think a lot of confusion comes from "one thing having different meanings." As in the example of the text shuffler above, such meanings do not come from the information source. Rather, they would come from the observers' knowledge of the meanings of certain arrangements of symbols, which has arisen through history and presumably been taught to them. This is the interaction of information in the signal and previously received, stored information in the individual, which requires computation. So, the Roman numerals VII doesn't store information about the number 7 explicitly itself, but rather it does so in the context of that relationship already having been transmitted to the observer and stored internally.

From the perspective of quantitative theories of information, which are used to define computation in physical systems, the text shuffler can only tell us information about the nature of the algorithm and how it shuffels the text, leaving aside information about the physical aspects of the computer. It doesn't have information on how to cure cancer, alternate endings to War and Peace, although it can produce text that can be interpreted in that way. This does set up a potential Gettier Problem in information theory though, although I haven't seen anyone write about it.

Having information rest solely in the minds of observers seems at risk of becoming subjective idealism. The information has to correspond to and emerge from external state differences or else how can we discuss incorrect interpretations of any signal?
• 853
Having information rest solely in the minds of observers seems at risk of becoming subjective idealism. The information has to correspond to and emerge from external state differences or else how can we discuss incorrect interpretations of any signal?

That's why i suggested "two player" game semantics. The semantics of interaction isn't accommodated by the traditional conceptions of either computation or causality, both which define life to be a one-player game but disagree as to who the solitary player is.
• 1.3k
I would argue that information exists "in the wild," as discernible differences.

Doesn't that imply that a discerner is a necessary condition for "discernible differences"? Or do you mean there are differences that are, potentially, discernible?
• 2.9k
how computation is instantiated in the world. . . . . Computation is what defines mathematical/abstract objects rather than it being some activity that you do with them.
I do have a theory of how "computation is instantiated in the world". But first, I must take issue with "computation" as a Definition rather than an Action*1. If you can accept -- as a philosophical postulation -- the notion that Evolution is a process of Computation (a la Tegmark), then my own unorthodox thesis might make sense.

It begins from the assumption that everything in this world is a form of Generic Information (Energy + Logic). The mathematical Logic of Nature gives direction to the propulsion of Energy. If so, then we can use a neologism to label that creative Enforming process : EnFormAction*2. I won't try to explain that novel concept further, unless you think that it could be a viable answer to your topical question : natural computation is instantiated via En-Form-Action -- the act of evolutionary computation of novel forms of being from previous entities. :smile:

*1. Computation : the action of mathematical calculation.
___Oxford
Note -- calculation adds or multiplies two or more values in order to derive a third value. Metaphorically, that's also what Evolution does, as it creates novel forms of being.

*2. EnFormAction :
That neologism is an analysis and re-synthesis of the common word for the latent power of mental contents : “Information”. “En” stands for energy, the physical power to cause change; “Form” refers to Platonic Ideals that become real; “Action” is the meta-physical power of transformation, as exemplified in the amazing metamorphoses of physics, whereby one kind of thing becomes a new kind of thing, with novel properties. In the Enformationism worldview, EnFormAction is eternal creative potential in action : it's how creation-via-evolution works.
https://bothandblog3.enformationism.info/page23.html
• 768

Doesn't that imply that a discerner is a necessary condition for "discernible differences"?

Yes, but not a conscious observer. For example, an indivisible "particle" alone in its own universe would transmit no information, and since it has no proper parts, no information transfer occurs within it. It cannot be interacted with. Can such a thing be said to exist? It would have no existence outside of some bare haecceity proposed as unobservable brute fact.

Scott Mueller's Asymmetry: The Foundation of Information has some good examples of relative indiscernibility in physical systems. We can usefully distinguish between "all possible discernable differences," and "all possible discernible differences vis-á-vis one systems interactions with its enviornment."

For the purposes of modeling physical systems, you can ignore "possibilities" that aren't relevant, but philosophically they seem relevant. Most people would like to avoid saying that differences go in and out of existence depending on what the system is interacting with. I think the idea that information only exists in the context of conscious observers is just a more specialized version of that unappealing view.
• 768

I will have to give this one some more thought.
• 17.4k
a riverbed wouldn't store information of the passage of water, but then its physical state, which seems identical to the total information that can be taken from it, is somehow different?

Tree rings contain evidence of forest fires, ice-cores atmospheric records. I'm not disputing that. But I'm saying that the mere existence of those data don't constitute information about anything until they're interpreted. The contrast to living organisms is that in them, information is dynamically interpreted by cellular processes moment by moment, it's intrinsic to any organic process.

Having information rest solely in the minds of observers seems at risk of becoming subjective idealism. The information has to correspond to and emerge from external state differences or else how can we discuss incorrect interpretations of any signal?

I think you're referring to a rather simplistic conception of idealism, of the variety that Samuel Johnson attempted to refute by kicking the stone. I favour a form of objective idealism. It's not that 'the world exists in my mind', but that what we understand as reality entails an ineliminable subjective aspect, without which nothing would make any sense. And we supply that. The mind is continually interpreting and integrating information about the world so as to make it intelligible - and not only intelligible, but navigable - for us. That order is at once 'the order of perceptions' and 'the order of the world' - in very much a Kantian sense.

All due respect, I think the error you're making is that of metaphysical naturalism - the assumption that the world would exist, just as it seems to now, were no humans present within it. But even that apparently empty world is still organised around an implicit perspective. Take that away and you can't imagine anything whatever.
• 1.3k
The contrast to living organisms is that in them, information is dynamically interpreted by cellular processes moment by moment, it's intrinsic to any organic process.

Wouldn't the interpretation have to be done by something with a mind?
• 17.4k
Not 'mind' in the sense of 'conscious mind'. It is purely cellular and organic in nature, but it still can be conceptualised in terms of interpretation of signs, hence, biosemiosis. (Don't want to introduce that as a subject of discussion however, just footnote.)
• 768

All due respect, I think the error you're making is that of metaphysical naturalism - the assumption that the world would exist, just as it seems to now, were no humans present within it. But even that apparently empty world is still organised around an implicit perspective. Take that away and you can't imagine anything whatever.

I don't make that mistake though. Without life, there is no color, no texture as such, perhaps no space-time as we understand it.

My objection is to the idea that fundemental differences in external objects somehow do not exist or change within the object when conscious observation occurs. I think the mechanisms which allow topsoil to record the passage of water, or passing light to record the existence of far off exoplanets, are the same mechanisms that allow eyes or cameras to record light, that the same mechanisms that make rocks vibrate due to pressure waves are involved in hearing, etc.

I don't want to get into the hard problem of consciousness, but simply the means by which sensory organs can record incoming data and neurons can subject that data to computation aren't qualitatively different from other natural phenomena.

So, my objection is to differences, of which information is composed, not existing simpliciter in external states. If mind is required to create them, then how to minds come to agree so much on that information? Why posit external objects at all if the fundemental source of all knowledge of them is only created by conscious observation?

Our differences might be on definition. I see information arising from fundemental ontological difference. Although I don't much like Floridi's overall theory of information, I find his arguments against popular conceptions of digital ontology in physics quite compelling (Chapter 14 of the Philosophy of Information). Quite simply, a universe without difference is impossible. Even a universe consisting of a two dimensional plane must have points whose coordinates differ from one another.

An ontology of fundemental difference is maximally "portable," in that it can fit with many other ontologies, be it flavors of idealism, dualism, or physicalism.

I would, however, agree with the physicists who push digital ontology on the idea that information is ontologically more basic than physical structures. These fundemental differences are a necessary condition for physical structures to arise. And in any event the "physical structures" we understand we only know as abstractions of mind, so in a both an ontological and epistemological sense, information is prior to physical state differences, not something that emerges from an interaction of mind and physical systems.

In my computation = cause thesis, which I am not very committed to, elementary elements of physics would be akin to numbers in formalist interpretations of Peano Arithmetic, while the more essential logic and relations are informational in nature. The axioms define the numbers, just as, in a universe with different constants, an electron would not be an electron and would behave differently. If I wanted to be even more speculative, I would say these "axioms" in physics are unlikely to be arbitrary brute facts existing as seemingly eternal laws, but rather the result of dialectical processes through which contradictions are resolved, and that this might explain the presence of mind in a teleological sense (sort of what Nagel has in mind for a project in his Mind and Cosmos). This is very speculative though, something like Basarab Nicolescu's book on Jacob Boehme and modern physics.

Perhaps our disagreement is on definition though. Semantic information or "meaning" appear to require mind and often times this is taken as synonymous with information, while I prefer the bare mathematical definition.
• 17.4k
Gee you have an interesting reading list! On mine is a book I might have mentioned previously, Mind and the Cosmic Order: How the Mind Creates the Features & Structure of All Things, and Why this Insight Transforms Physics, by Charles Pinter. That book condensed many of my ideas about the foundational role of the mind. It’s more about neurological modelling than philosophy as such but it has profound philosophical implications, I feel. (It went a bit under the radar, because Pinter is a maths emeritus, not a philosopher, and I don't think it got much notice in academia, but I thought it an important book.)

The axioms define the numbers, just as, in a universe with different constants, an electron would not be an electron and would behave differently.

You’re familiar with books such as ‘Just Six Numbers’ by Martin Rees? (Achingly dull read, I found.) It's about the fundamental physical constraints which must exist at a foundational level if the universe is even going to form matter. So I don't know if it's feasible that there could be an alternative, there's something about necessity woven into the fabric of the cosmos, seems to me. These ratios and values have to be a certain way, otherwise stars would not form.

As for information - I think the difference we have is roughly like the difference between pan- and biosemiosis. Pansemiosis proposes that all things, living and non-living, possess a form of semiotic or sign-making capacity, that everything in the universe, including animals, plants, rocks, and even inanimate objects, can be interpreted in terms of signs. Biosemiosis limits the scope of semiotics to living processes. It's an area of disagreement, but the latter seems more feasible to me.
bold
italic
underline
strike
code
quote
ulist
image
url
mention
reveal