Tuesday, May 8, 2012

Kripke contra computationalism

That the brain is a digital computer and the mind the software run on the computer are theses that seem to many to be confirmed by our best science, or at least by our best science fiction.  But we recently looked at some arguments from Karl Popper, John Searle, and others that expose serious (indeed, I would say fatal) difficulties with the computer model of the mind.  Saul Kripke presents another such argument.  It is not well known.  It was hinted at in a footnote in his famous book Wittgenstein on Rules and Private Language (WRPL) and developed in some unpublished lectures.  But Jeff Buechner’s recent article “Not Even Computing Machines Can Follow Rules: Kripke’s Critique of Functionalism” offers a very useful exposition of Kripke’s argument.  (You can find Buechner’s article in Alan Berger’s anthology Saul Kripke.)

Though it is, I think, not essential to Kripke’s argument, the “quus” paradox developed in WRPL provides a helpful way of stating it (and, naturally, is made use of by Kripke in stating it in WRPL).  So let’s briefly take a look at that.  Imagine you have never computed any numbers as high as 57, but are asked to compute “68 + 57.”  Naturally, you answer “125,” confident that this is the arithmetically correct answer, but confident also that it accords with the way you have always used “plus” in the past, i.e. to denote the addition function, which, when applied to the numbers you call “68” and “57,” yields 125.  But now, Kripke says, suppose that an odd skeptic asks you how you are so sure that this is really what you meant in the past, and thus how you can be certain that “125” is really the correct answer.  Maybe, he suggests, the function you really meant in the past by “plus” and “+” was not addition, but rather what Kripke calls the “quus” function, which he defines as follows:

x quus y = x + y, if x, y < 57;
              = 5 otherwise.

So, maybe you have always been carrying out “quaddition” rather than addition, since quadding and adding will always yield the same result when the numbers are smaller than 57.  That means that now that you are computing “68 + 57,” the correct answer should be “5” rather than “125.”  And maybe you think otherwise only because you are now misinterpreting all your previous uses of “plus.”  Of course, this seems preposterous.  But how do you know the skeptic is wrong?

Kripke’s skeptic holds that any evidence you have that what you always meant was addition is evidence that is consistent with your really having meant quaddition.  For example, it is no good to note that you have always said “Two plus two equals four” and never “Two quus two equals four,” because what is in question is what you meant by “plus.”  Perhaps, the skeptic says, every time you said “plus” you meant “quus,” and every time you said “addition” you meant “quaddition.”  Neither will it help to appeal to memories of what was consciously going through your mind when you said things like “Two plus two equals four.”  Even if the words “I mean plus by ‘plus,’ and not ‘quus’!” had passed through your mind, that would only raise the question of what you meant by that

Note that it is irrelevant that most of us have in fact computed numbers higher than 57.  For any given person, there is always some number, even if an extremely large one, equal to or higher than which he has never calculated, and the skeptic can always run the argument using that number instead.  Notice also that the point can be made about what you mean now by “plus.”  For all of your current linguistic behavior and the words you are now consciously running through your mind, the skeptic can ask whether you mean by it addition or quaddition. 

Now, Kripke’s “quus” puzzle famously raises all sorts of questions in the philosophy of language and philosophy of mind.  This is not the place to get into all that, and Kripke’s argument against functionalism does not, I think, stand or fall with any particular view about what his “quus” paradox ultimately tells us about human thought and language.  The point for our purposes is that the “quus” example provides a useful illustration of how material processes can be indeterminate between different functions.  (An Aristotelian-Thomistic philosopher like myself, by the way, is happy to allow that mental imagery -- such as the entertaining of visual or auditory mental images of words like “plus” or sentences like “I mean plus, not quus!” -- is as material as bodily behavior is.  From an A-T point of view, among the various activities often classified by contemporary philosophers as “mental,” it is only intellectual activity in the strict sense -- activity that involves the grasp of abstract concepts, and is irreducible to the entertaining of mental images -- that is immaterial.  And that is crucial to understanding how an A-T philosopher would approach Kripke’s argument.  But again, that is a topic for another time.)

Kripke’s “quus” example can be used to state his argument about computationalism as follows.  Whatever we say about what we mean when we use “plus,” there are no physical features of a computer that can determine whether it is carrying out addition or quaddition, no matter how far we extend its outputs.  No matter what the past behavior of a machine has been, we can always suppose that its next output -- “5,” say, when calculating numbers larger than any it has calculated before -- might show that it is carrying out something like quaddition rather than addition.  Of course, it might be said in response that if this happens, that would just show that the machine was malfunctioning rather than performing quaddition.  But Kripke points out that whether some output counts as a malfunction itself depends on what program the machine is running, and whether the machine is running the program for addition rather than quaddition is precisely what is in question.  

Another way to put the point is that the question of what program a machine is running always involves idealization.  In any actual machine, gears get stuck, components melt, and in myriad other ways the machine fails perfectly to instantiate the program we say it is running.  But there is nothing in the physical features or operations of the machine themselves that tells us that it has failed perfectly to instantiate its idealized program.  For relative to an eccentric program, even a machine with a stuck gear or melted component could be doing exactly what it is supposed to be doing, and a gear that doesn’t stick or a component that doesn’t melt could count as malfunctioning.  Hence there is nothing in the behavior of a computer, considered by itself, that can tell us whether its giving “125” in response to “What is 68 + 57?” counts as an instance of its following an idealized program for addition, or instead as a malfunction in a machine that is supposed to be carrying out an idealized program for quaddition.  And there is nothing in the behavior of a computer, considered by itself, that could tell us whether giving “5” in response to “What is 68 + 57?” counts as a malfunction in a machine that is supposed to be carrying out an idealized program for addition, or instead as an instance of properly following an idealized program for quaddition.

As Buechner points out, it is no good to appeal to counterfactuals to try to get around the problem -- to claim, for example, that what the machine would have done had it not malfunctioned is answer “125” rather than “5.”  For such a counterfactual presupposes that the idealized program the machine is instantiating is addition rather than quaddition, which is precisely what is in question. 

Naturally, we could always ask the programmer of the machine what he had in mind.  But that simply reinforces the point that there is nothing in the physical properties of the machine itself that can tell us.  But if there is nothing intrinsic to computers in general that determines what programs they are running, neither is there anything intrinsic to the human brain specifically, considered as a kind of computer, that determines what program it is running (if it is running one in the first place).  Hence there can be no question of explaining the human mind in terms of programs running in the brain.

Might we appeal to God as the programmer of the brain who determines which program it is running?  Obviously most defenders of the computer model of the mind would not want to do this, since they tend to be materialists and materialists tend to be atheists.  But it is not a good idea in any case.  For that would make of human thought something as extrinsic to human beings as the program a computer is running is extrinsic to a computer, indeed as extrinsic as the meaning of a sentence is to the sentence.  Just as the meaning of “The cat is on the mat” is not really in the sounds, ink marks, or pixels in which the sentence is realized, but rather in the mind of the user or hearer of the sentence, so too the idea of God as a kind of programmer or user of the brain qua computer would entail that the meanings of our thought processes are not really in us at all but only in Him.  The result would be a new riff on occasionalism that is even more bizarre than the usual kind -- a version on which it is really God who is, strictly speaking, doing all our thinking for us!

Neither, as Buechner points out, will it do to suggest that natural selection has determined that we are following one program rather than another.  For any program we conjecture natural selection has put into us, there is going to be an alternative program with equal survival value, and the biological facts will be indeterminate between them.  There will be no reason in principle to hold that it is the one program that natural selection put into us rather than the other.

Suppose we say instead that there is what Buechner calls a “telos in Nature” that determines that the brain really is following this program rather than that -- the program for addition, say, rather than quaddition?  In that case we would have some end or purpose intrinsic to the natural world that determines which program the brain instantiates, which would eliminate the occasionalist problem the appeal to God as programmer raised.  (Of course, you could give a Fifth Way style argument for God as the ultimate explanation of this intrinsic telos, but that would not be to make of God a “programmer” in the relevant sense, any more than Aquinas’s Fifth Way makes of God a Paley-style tinkerer.)

Buechner himself is not sympathetic to this “telos in Nature” suggestion, but it is, naturally, one that an Aristotelian is bound to take seriously.  But it does not help the advocate of the computer model of the mind, at least not if he is a materialist.  For to affirm that there is teleology intrinsic in nature is just to abandon the materialist’s conception of matter and return to something like the Aristotelian-Scholastic conception that materialists, like other modern philosophers, thought they had buried forever back in the days of Hobbes and Descartes.  

Still, if the computer model of the mind leads people to reconsider Aristotelianism, it can’t be all bad.  (Cf. James Ross’s “The Fate of the Analysts: Aristotle’s Revenge: Software Everywhere”)

283 comments:

  1. I'm not seeing Kripke's argument. Why can't the skeptic be answered by anyone who knows how 'plus' is defined based on Peano's axioms? Because I know that, I know that no matter high a number I am adding to another, it is not a quus operation. And once a computer is programmed with a plus function of that nature (for that matter, it could be built into the hardware), then that definition of plus, which will give 125, not 5, can be shown to the skeptic to be a physical property of the machine.

    ReplyDelete
  2. Whatever we say about what we mean when we use “plus,” there are no physical features of a computer that can determine whether it is carrying out addition or quaddition, no matter how far we extend its outputs.

    Maybe you can't, but anyone who understands the workings of the computer can easily inspect it to understand what it is doing and what it will do on inputs it hasn't seen yet.

    Another way to put the point is that the question of what program a machine is running always involves idealization. In any actual machine, gears get stuck, components melt, and in myriad other ways the machine fails perfectly to instantiate the program we say it is running.

    Well that is accurate at least.

    But there is nothing in the physical features or operations of the machine themselves that tells us that it has failed perfectly to instantiate its idealized program.

    Consider: "there is nothing in the physical features of a cow that tells us whether it is fulfilling its functions or is just one sick, malfunctioning, or possibly dead cow."

    This is pretty obviously false, so why should the similar statement about computers be taken to be obviously true?

    ReplyDelete
  3. My question is more or less the same as SR's: Why wouldn't invoking the axioms of Peano arithmetic be an absolute rebuttal to the skeptic's argument?

    ReplyDelete
  4. I suggest swapping "Schmeano's axioms" for Peano's as Kripke swaps quus for plus and pondering the argument afresh. Grueness is a related problem and James Ross discusses the issues very well in "Immaterial Aspects of Thought". A large discussion of Ross's argument can be found by searching this blog or my veniaminov.blogspot.com blog. Same immediate objections always come up so maybe reviewing a previous discussion will spare a reinvention of the dialectical wheel.

    ReplyDelete
  5. As GIP say, in principle by materialist reduction, the exact process of plus being used can be determined. However, in terms of minds I don't see why any of this matters since I agree this is just how minds work. Discrepancies only arise when different assertions conflict, like between 5 and 125. When there is no conflict, there is no problem. That seems to be exactly how exchanges between minds work. We are separate vessels passing messages to one another, and it does come as a shock from time to time when one of us realizes we meant 'plus' while another something subtly or quite distinctly different.

    ReplyDelete
  6. As GIP say, in principle by materialist reduction, the exact process of plus being used can be determined. However, in terms of minds I don't see why any of this matters since I agree this is just how minds work. Discrepancies only arise when different assertions conflict, like between 5 and 125. When there is no conflict, there is no problem. That seems to be exactly how exchanges between minds work. We are separate vessels passing messages to one another, and it does come as a shock from time to time when one of us realizes we meant 'plus' while another something subtly or quite distinctly different.

    ReplyDelete
  7. @Codgitator,

    Thanks for the pointer, but I am still confused. I agree with Ross that a machine is not adding in the formal sense, but I still do not see what plus/quus has to do with it. I would think that one cuold answer the skeptic with "I know I am adding, and not quadding, because I am following an ideal process for doing so." And one can tell whether or not a computer is (simulating) adding rather than quadding by looking at its wiring. That's where I get lost, when the post says

    there is nothing in the physical properties of the machine itself that can tell us [whether it is adding or quadding].

    I would think that would only be the case if one could only look at its inputs and outputs.

    ReplyDelete
  8. GIP,

    "Maybe you can't, but anyone who understands the workings of the computer can easily inspect it to understand what it is doing and what it will do on inputs it hasn't seen yet."

    Let 0=high voltage and 1=low voltage

    A certain simple computer has two inputs corresponding to 0101 and 1010. You inspect the hardware and determine that the output will be 1111. What is the computer doing?

    ReplyDelete
  9. To extend David's thought: I have a Texas Instruments TI-25. When I put in a series of entries that *to some observers* might look like I was trying to add 5.3x10^66 plus 6.7x10^93, the output is "Error". Who is to say whether the machine's design is to perform the querraddition operation, which is like addition below 10^99 and returns "Error" above 10^99, or whether it does the "plus" operation below 10^99, and the "querror" operation above, or whether it is only designed to do the plus operation and is FAILING that operation above 10^99? Or, rather, who is to determine whether the machine is designed to "perform the plus operation when dry and to blow circuits when submerged in water"? What's that, you say "nobody designs a machine to blow circuits"? Well, sorry, but that's exactly what a fuse is designed to do.

    For a machine that is inspected and appears to construct a mechanical algorithm to follow Peano's axioms for addition, who is to say that the machine's proper input is 2 numbers instead of having an anvil dropped on it?

    ReplyDelete
  10. Hi,

    I've just started looking at this blog over the last few days - good stuff. I'm a research scientist, a Roman Catholic, and I have a PhD in Computer Science.

    I find this post rather vague. The assumptions about what is observable, and how the computer might be modelled, are not made clear at the outset.

    There are two obvious ways in you could model the computer. One is as a Turing Machine (a rule box and a read/write tape); the other is as an actual practical machine which might have a standard set of machine instructions, an operating system, and application programs. The latter model might be particularly relevant here, because certainly in terms of basic machine instructions all computers quadd rather than add - there is always a maximum value above which they cannot operate. This is an inherent limitation of say 32-bit or 64-bit architectures. If you want to do better than that and perform pseudo-infinite precision arithmetic, you need higher level application software to compensate for the limitations of the machine instructions. But ultimately all practical machines will run out of bits. All quadd.

    The three level layer is also interesting because they are typically provided by different interests - Intel chooses the machine instruction set, Microsoft or Apple chooses the operating system and its interface, and while anyone can write the application software, increasingly these are provided by yet other major commercial interests and users restrict themselves to very sophisticated configuration of the applications they use, plus providing vast amounts of data. So who owns the meaning? This might vary from layer to layer, and a key to success is always to enable your customer to create higher level stuff that your system facilitates as painlessly as possible.

    The same three layers could very crudely apply to the neuron (basic machine instruction set?) unconscious (Operating system?) and applications (conscious activities?).

    But the most pertinent questions apply to observation. The post talks of 'seeing' the operation of the computer. Is that the flashing lights on the outside? The output on the screen? The voltage levels inside the full memory and the processor registers? Once we get to this level, then perhaps we can reverse engineer the intention of the programming, if we can guess the instruction set. Or maybe we get the manual and the application code itself? The more we can see, the more we can interpret and predict behaviour, assuming the system has been built rationally.

    ReplyDelete
  11. @goddinpotty:

    "Maybe you can't, but anyone who understands the workings of the computer can easily inspect it to understand what it is doing and what it will do on inputs it hasn't seen yet."

    Just to spell out David T.'s point, there are an infinite possible number of functions that are instantiated by whatever it is that the computer is doing, so even armed with all your knowledge of the inner workings of a computer, alongside your misunderstanding of the argument, you can't either.

    Human thinking *must* be determinate in such a precise way that no function among physical states can be -- in a nutshell, this is precisely what Kripke's argument shows. They have to exclude all the incompossible forms for otherwise they would not have the features that characterize them, but since whatever it is that the computer is doing *cannot* rule out such incompossible forms, it can never be determinate in way that allows us to say that the computer *is* doing addition. Or to use James Ross' phrasing

    The single case of thinking has to be of an abstract "form" (a "pure" function) that is not indeterminate among incompossible ones. For instance, if I square a number-not just happen in the course of adding to write down a sum that is the square, but if I actually square the number-I think in the form "N
    X N = N2."


    You can replace addition, by an application of modus ponens or a valid application of the infinitary axiom schema of induction of first order Peano arithmetic (this last example is for those who have invoked the Peano's axioms).

    ReplyDelete
  12. Hi Ed,

    Thanks for your thought-provoking post. I used to be a computer programmer for ten years, so I have to say that the statement that "there is nothing in the physical properties of the machine itself that can tell us [whether it is adding or quadding]" is simply mistaken.

    If you want a good, clear explanation of how a computer adds, check out this Web site:

    http://www.physicsforums.com/
    showthread.php?t=18870

    and scroll down to Chen's comment at Apr10-04, 10:04 AM.

    It isn't a matter of inputs and outputs - I agree they are ambiguous. Logic gates are not. Hence grodrigues' comment that "there are an infinite possible number of functions that are instantiated by whatever it is that the computer is doing" misses the point. There's more to a computer than what it actually does (functionality). There's also what it can and can't do, which is determined by its architecture. This answers Tony's comment as well. If we ever discovered a calculator on an alien planet, we'd certainly be able to figure out if it was an adding or a quadding machine.

    That was all I wanted to say.

    ReplyDelete
  13. I am not sure that it is safe to say, as goddinpotty does that "we can easily inspect it to understand what it is doing" or as Vincent Torley said "If we ever discovered a calculator on an alien planet, we'd certainly be able to figure out if it was an adding or a quadding machine."

    This is reasonable enough for the simple case of adding vs. quadding, but in the general case one cannot figure out what a formal system does by inspecting it, because that can be reduced to the halting problem. The computer may be a human being (question begged) who is trying to figure out what we are doing, and from this we could construct a paradox.

    Also amusing for the purposes of this example is that inspecting a typical computer will indicated clearly that it is not adding. Peano would not allow us to return to our starting place by adding 1 enough times, but the computer may overflow it's registers. The add operation is that of a cyclic group.

    ReplyDelete
  14. @Vincent Torley:

    With all due respect, it is you who are missing the point. Having had to endure some dreadfully boring courses on digital systems, and even having toyed with classics such as Henessy and Patterson's "Computer Architecture" on computer hardware, I do know a little bit about how computers work and can say with some confidence that such knowledge is just *irrelevant* to the argument.

    Yes, a logic gate, say an AND or XOR, performs a specific function (sets some bits, say in a register, to 0 or 1), and concatenating a number of such logic gates we can implement the high-school algorithm of addition for numbers encoded in binary form. But even if we grant for the sake of argument that a computer has, or could have, an unbounded memory, it is still a matter of objective fact that whatever it is that it does, it instantiates an infinite number of incompossible functions so it cannot be determinate in the way human thought is. It can implement addition, or some bizarre function on strings of characters (fix whatever encoding you care, ascii, utf16, whatever) or basically anything, if we encode the inputs just right.

    ReplyDelete
  15. @DavidT -- you are conflating two issues.

    1) If you just look at the inputs and outputs, you cannot tell for certain what the computer is doing (you can apply abductive reasoning to guess that it will always add its inputs, but that is not certain). But if you inspect the internal mechanism, you can.

    2) Your other point seems to be about the interpretation of inputs and outputs. For that, imagine you have a dog that is barking, and you its owner need to interpret what it means (hunger? an intruder?). Or imagine you are a heartless scientist tracing the neural path between the dogs perception and effectors to find the mechanical causal path between the stimulus and response. Both are perfectly plausible scenarios, because the dog is both a mechanical system and an intentional system. And so is a a computer.

    ReplyDelete
  16. I'm really amazed at how badly Vincent and GIP are missing the point. Yes, given knowledge of computer science etc., we know that such-and-such is a logic gate, that the machine is adding, etc. But that (rather obviously) presupposes that the machine was designed by people who themselves had knowledge of computer science, for the purpose of instantiating such and such a program. Kripke's point is that you are not in principle going to be able to determine what the machine is doing in the absence of such background knowledge; that is to say, you are not going to read it off from those properties of the machine that can be defined entirely in terms of physics, with no reference to the intentions of the designers.

    So, to refute Kripke, you have to find some property or properties of the machine which (a) make no reference whatsoever, not even implicitly, to the intentions of people who design logic gates, write programs, etc. but still (b) suffice all by themselves to determine that the machine is running this program rather than that. Good luck.

    ReplyDelete
  17. Frankly, I dont get the "quus-paradox". Addition, as I see it, falls under the category of action. And a program is just a certain type of complex action, analysed into actions of a simpler type. (One should distinguish the programm and its description, i.e. the code in which it is expressed)

    Now, every aspect of reality is completly determined, i.e. every substance or accidens instantiates a given universal or not. (The Law of Noncontradiction, which implies the Law of Excluded Middle, is not a law of thought, but a law of reality. If it were a law of thought, reductio-ad-absurdum arguments would be unthinkable, since they involve a contradictory assumption) Therefore it is completly determined, whether a complex action of a machine instantiates a program or not.

    Or, are we talking about whether a given rational agent is able to determine if the computations of a machine instantiate a certain type of action or not? Well, even if he couldnt, what would be the point of that?

    ReplyDelete
  18. Define a function f as
    int f(int a, int b) {
    return a+b;
    }

    Now, is f a function that adds two numbers together and returns the result? Sure, under one possible interpretation of f. The key point here is that this interpretation is not the only one you can make for f instantiates other possible interpretations (uses) as well.

    Int here is not necessarily a number. It can be a letter, a vertex property, a state indicator etc. Peanos axioms require that a successor function S(x) exists for a or b, without having defined a successor function the addition function breaks down. Certainly the interpretation of an int fits the Successor function definition (up to a point but whatever) since there are successors to any numbers (3 follows 2, 4 follows 3 etc), it also, however fits many other sorts of things and there is no reason to deduce that f here is in fact adding two numbers for a and b might not have a successor function, they might not be numbers, the + sign in that case does not fit the Peanos axioms description for addition and is instead something else.

    So, an example:

    string encryptMessage(string oldMessage) {
    string newMessage = "";
    for_each(int letter in oldMessage) {
    int encryptionKey = (int from binary)0100111010101;
    int encryptedLetter = f(letter, encryptionKey)
    newMessage.put(encryptedLetter);
    }
    return newMessage;
    }

    Here we are encrypting a message by running the function f but not adding two numbers together (although that is a possible interpretation if one does not know what a and b represent). We can know that we are not adding because a letter and our encryptionKey does not have a Successor function.

    ReplyDelete
  19. GIP,

    I don't think I am conflating anything. Assume whatever you want about the hardware, and as detailed a knowledge as you like about it. How about this: 1010 and 0101 are both inputs to a piece of hardware, conventionally known as an adder, with 1111 as output. What has happened here?

    ReplyDelete
  20. @Feser: Yes, if you approach a complex artifact or lifeform with no knowledge whatsover it is hard, probably impossible, to figure out what it is doing (or supposed to be doing). This is a staple SF scenario in fact, humans confronted by an inexplicable alien artifact.

    However, this is not an interesting scenario because it is counterfactual. In reality, we do have background knowledge, we share common origins and a common world with the intentional artifacts and lifeforms we are trying to understand.. So Kripke's point, insofar as I understand it, is simply irrelevant and uninteresting (a reaction I have to all of Kripke's work -- I know he is supposed to be a big noise, but I just do not get it).

    So, to refute Kripke, you have to find some property or properties of the machine which (a) make no reference whatsoever, not even implicitly, to the intentions of people who design logic gates, write programs, etc. but still (b) suffice all by themselves to determine that the machine is running this program rather than that. Good luck.

    As indicated above, I have no interest in refuting Kripke, because his argument is irrelevant to the real world, which is what I care about.

    ReplyDelete
  21. Vincent,

    You surely get the point I am making to GIP. 1010 + 0101 = 1111 can be interpreted as 10+5=15. Or it can equally be interpreted as -6+5=-1. The operation is physically the same in the computer but has different meaning depending on the intent of the programmer. This is what makes computers so powerful as "universal machines." The physically identical operation is open to multiple interpretations. The same principle follows all the way up the software hierarchy. Modular software is physically identical yet is capable virtually unlimited interpretations depending on the needs of an engineer. Absent the engineer, it doesn't mean anything at all, just as 1010 + 0101 doesn't mean anything absent someone to give it an interpretation.

    ReplyDelete
  22. Yes, if you approach a complex artifact or lifeform with no knowledge whatsover

    You're attacking a straw man. I didn't say "no knowledge whatsoever." I said "no knowledge of the intentions of the designers." You can have all the other knowledge you want. And while your straw man is trivial, Kripke's actual argument is not.

    ReplyDelete
  23. OK, so limit the discussion to knowledge of the intentions of the designers, or in the case of living systems, of the constraints and functionality imposed by natural selection. Don't see how that changes the argument. My point is that you do have such knowledge so Kripke's argument only applies to an irrelevant case.

    ReplyDelete
  24. @goddinpotty:

    "Don't see how that changes the argument. My point is that you do have such knowledge so Kripke's argument only applies to an irrelevant case."

    The first two words aptly summarize your stance. Kripke's argument shows -- and you seem to concede its validity -- that any explanation of the mind rejecting intentionality cannot succeed, not even in principle, and therefore, the computational explanation of the mind, or more generally naturalistic explanations of the mind, which do evacuate intentionality to the dustbin of irrelevance, cannot succeed, not even in principle. Ergo, computational or naturalistic explanations are of necessity incomplete. Do I need to go on and spell out what this means?

    ReplyDelete
  25. External reality is the common reference point between minds, and brains, the hardware of minds, are general pattern recognition devices. When two people discuss a bird in a tree, how do they know that their minds are actually representing the same thing? Or, let's say we consider the possibility of ever discussing anything with a robot. How will there be any congruence between our thoughts when the substrate hardware will be radically different? The answer is that we share a common external reality and the patterns our brains have learned correlate at an abstract level. There needn't be a one to one, bit to bit or synapse to synapse mapping.
    However, when even the abstract mapping breaks down, like when you are discussing someone's wife and they think you're discussing a hat, a conflict arises that suggests the incongruity of your thoughts.

    ReplyDelete
  26. @grodriguez: you seem to think there is some disjunction between "naturalistic explanations" and "intentionality". As I've explained in earlier comment threads here, I find that entirely bogus. Intentionality is as natural as anything else.

    Of course if you don't believe that, and instead think that intentionality or purpose can only be supplied by sprinkling supernatural fairy dust over a natural system, well, that is the crux of the biscuit. This is basically the argument from lack of imagination, and I don't think that Kripke adds anything to it.

    ReplyDelete
  27. GIP,

    My point is that you do have such knowledge so Kripke's argument only applies to an irrelevant case.

    Compare: Suppose you pointed out to someone that the meaning of the sentence "The cat is on the mat" is entirely conventional, and does not derive from the physical properties of the symbols -- their shape, the chemistry of the ink in which they are written, etc. To underline the point, suppose you note that someone who knew only such physical properties but knew nothing about the English language would not be able to determine the meaning of the sentence.

    Now suppose that someone replied to this by saying: "Well, we do in fact know about English, what the sentence means in English, etc. So this is an irrelevant case and proves nothing." Would this be a good reply? Obviously not -- such a person would simply be missing the whole point.

    Same with you. Your reply to Kripke is no less clueless.

    Of course if you don't believe that, and instead think that intentionality or purpose can only be supplied by sprinkling supernatural fairy dust over a natural system, well, that is the crux of the biscuit.

    How exactly does Kripke's argument depend on belief in "supernatural fairy dust" or any other such New Atheist style straw man? In fact the argument presents no positive theory at all, but only a critique of a certain specific naturalist theory. Let intentionality be as physical as you wish, that wouldn't show in the least that the computer model of the mind, in particular, is correct.

    This is basically the argument from lack of imagination,

    Where does Kripke say or imply anything that amounts to "I can't imagine how this is true, therefore it's not?"

    Again, suppose your hypothetical interlocutor said "Your claim that the meaning of the sentence 'The cat is on the mat' is conventional just shows that you have mistaken your failure of imagination for an argument." Would that be a clever response? Obviously not. And neither is your response to Kripke.

    ReplyDelete
  28. I think my (inexpert) solution and what GIP is implying is Kripke's skeptical solution. The paradox applies to meaning for individuals in isolation, and the skeptical solution grounds meaning in community. Whether a computing machine in isolation "means" to do plus or quus becomes irrelevant when it becomes a component part of a community of machines. In that situation, the only relevant thing is that it performs the operation correctly. Analogously, the same thing can be said about what you or I mean by "cat" or "hat" in isolation as compared to in community with others. I prefer to express it in terms of message passing and conflict in communication, but the idea is the same. In isolation there is no significance to the symbols used, but in community they must correspond to avoid conflict.

    ReplyDelete
  29. Compare: Suppose you pointed out to someone that the meaning of the sentence "The cat is on the mat" is entirely conventional, and does not derive from the physical properties of the symbols -- their shape, the chemistry of the ink in which they are written, etc. To underline the point, suppose you note that someone who knew only such physical properties but knew nothing about the English language would not be able to determine the meaning of the sentence.

    You are trying to pull the same trick that grodrigues does (not that I think it's trickery on your part or his, I'm sure it is based on sincere belief -- but to me, it appears to be an attempt to sneak in the same invalid assumption through a slightly different door). Namely, setting up a disjunction between the "physical" and the "conventional".

    The best response to this is the example of genetic translation that I have offered up here before. The mapping from DNA triplets to amino acids is entirely "conventional" in the sense that it could easily be different. But it is also "physical", because the mapping is implemented by a population of tRNA molecules that bind to a segment of mRNA and a particular amino acid. Indeed, the translation conventions vary slightly in some organisms and in mitochondria.

    In your example, yes the relation between "cat" and real cats is "conventional", but that convention is implemented by physically embodied brains that happen to share a language. Conventionality is not something separate from physicality.


    How exactly does Kripke's argument depend on belief in "supernatural fairy dust"...Where does Kripke say or imply anything that amounts to "I can't imagine how this is true, therefore it's not?"

    That was in reference to grodrigues' built-in assumption that intentionality and the natural are disjoint.

    Again, suppose your hypothetical interlocutor said "Your claim that the meaning of the sentence 'The cat is on the mat' is conventional just shows that you have mistaken your failure of imagination for an argument." Would that be a clever response? Obviously not. And neither is your response to Kripke.

    See above. I can imagine how meaning is physically implemented, because I've spent a long time working with both computers and biology. Most people haven't, and so they do suffer from "lack of imagination".

    I will apologize for being snarky; it is no crime to not to have my particular background. What is a crime (or at least, annoying) is willfully persisting in insisting that something is impossible in the face of good examples of how it isn't.

    ReplyDelete
  30. @goddinpotty:

    "That was in reference to grodrigues' built-in assumption that intentionality and the natural are disjoint."

    It is not my assumption, it is the contention of virtually all metaphysical naturalists that teleology is not immanent, but rather ultimately reducible to the mechanistic operation of physical constituents. You disagree? Welcome to the fold.

    "I can imagine how meaning is physically implemented, because I've spent a long time working with both computers and biology. Most people haven't, and so they do suffer from "lack of imagination".

    I will apologize for being snarky; it is no crime to not to have my particular background. What is a crime (or at least, annoying) is willfully persisting in insisting that something is impossible in the face of good examples of how it isn't."

    And what is especially annoying is imagining that the arcane knowledge you possess about computers and biology, an arcane knowledge shared by millions of people in the world, is somehow relevant, with the implied assumption that only if everybody here knew what you know we would immediately see the sillyness of say, Kripke's argument. Guess what, not everybody here is as clueless about computers as you are of the matter being discussed. It was already explained why such knowledge, no doubt admirable, will not help you one iota in refuting Kripke.

    ReplyDelete
  31. GIP,

    I will apologize for being snarky; it is no crime to not to have my particular background. What is a crime (or at least, annoying) is willfully persisting in insisting that something is impossible in the face of good examples of how it isn't.

    We've all seen your examples. We've all pointed out the flaws in both the examples and your reasoning about them. Repeatedly. Whether intentionally or not, you simply do not grasp the arguments being presented to you.

    I'd illustrate why, but holy hell - Ed, Grod and others (including at least one, likely more than one, computer programmer / compsci person) have done so, repeatedly, now and over the course of months. You haven't gotten it, and I think at this point the smart money says you're not going to, barring a eureka moment on your part. PEBKAC and all that.

    I offer this up mostly to explain what I think is inevitably going to soon follow - you're going to lodge the same failed objection for the hundredth time, and people are just going to blow past it to engage questions and critiques from people who aren't subject to the same limitation.

    ReplyDelete
  32. Appropriately enough, the "Central Dogma" is a good example at the molecular level of how meaning works. Just floating on its own, an mRNA means nothing, it's the ribosomes, tRNA, aa's, and of course, the resulting active protein and molecular milieu of the cell (the overall community of molecules) that provide what meaning exists. It's not even an entirely "dumb" process, since it's a many to one (64 to 20) degenerate mapping.

    What Dr. Feser is doing is isolating an individual molecule and proclaiming that it can mean nothing (and he's right), while ignoring everything else.

    ReplyDelete
  33. "Namely, setting up a disjunction between the 'physical' and the 'conventional'."

    So . . . then you're a metaphysical realist?

    ReplyDelete
  34. "you simply do not grasp the arguments being presented to you."

    Don't worry. They say this to everyone who doesn't agree with them. They can't be wrong. You just don't understand.
    :)

    ReplyDelete
  35. Hi Hunt, yes, that seems to be the usual response. Disagreement can only be the result of failure to understand, or failure to read a couple thousand pages of Aquinas, or the like.

    ReplyDelete
  36. Well, either "they" don't understand or you don't understand. Which do you think is more likely?

    ReplyDelete
  37. t_t so cute, GIP has an ally now!

    o_o still...

    ReplyDelete
  38. @Hunt:

    "you simply do not grasp the arguments being presented to you.

    Don't worry. They say this to everyone who doesn't agree with them. They can't be wrong. You just don't understand."

    Funny, that is what the clueless ones usually say. They say it to everyone who disagrees with them. They can't be wrong. The others just don't understand.

    note: putting smileys at the end of a post to signal irony is one of the most contemptible violences that the technologies of distraction have done to written speech.

    ReplyDelete
  39. In his first comment above Ed says

    >> Kripke's point is that you are not in principle going to be able to determine what the machine is doing in the absence of such background knowledge; that is to say, you are not going to read it off from those properties of the machine that can be defined entirely in terms of physics, with no reference to the intentions of the designers. <<

    I don't agree. We can give a description of the behaviour of a physical system like an adder at a number of levels. We can look at its structure, apply our understanding of physics (easier in the case of a mechanical adding machine, perhaps) and conclude that the output must bear a certain relation to the inputs that can be conveniently described as follows: if the inputs can be described as the binary representations of natural numbers m and n, then the output can be described as the binary representation of the sum m+n, perhaps modulo some power of two. Is this not purely descriptive, with no reference to intentions?

    ReplyDelete
  40. Hunt, if you really understand Dr. Feser's argument, restate it in your own words. Then we'll know whether or not you really do. So far, it appears obvious you don't.

    ReplyDelete
  41. We can give a description of the behaviour of a physical system like an adder at a number of levels. We can look at its structure, apply our understanding of physics (easier in the case of a mechanical adding machine, perhaps) and conclude that the output must bear a certain relation to the inputs that can be conveniently described as follows: if the inputs can be described as the binary representations of natural numbers m and n, then the output can be described as the binary representation of the sum m+n, perhaps modulo some power of two. Is this not purely descriptive, with no reference to intentions?

    I think there's going to be numerous objections to this. For one, 'described as the binary representations' seems to involve doing what Feser/Kripke suggests is inevitable, and bringing in intentions. I have 3 empty cans of Sprite Zero on my desk, and a fan aimed at me. I suppose I could say that, if I consider the cans to each represent the number 3, then together they represent 9, and if one falls off the desk then that's a subtraction operation. But the representation is in my mind - not in the cans.

    Just to lodge a quick objection for the moment.

    ReplyDelete
  42. "Funny, that is what the clueless ones usually say. They say it to everyone who disagrees with them. They can't be wrong. The others just don't understand."

    Funny, that's what the really clueless ones say when I say it to them.

    :)
    ;)

    ReplyDelete
  43. I like the paradox and accept the solution offered, but I wonder if the solution does not offer an even bigger skeptical problem than the original skeptic.

    I take it that the conclusion can be generally stated thus: Unless we appeal to some thing outside the system in question, we can never determine if the output of the system is the accurate output of an algorithm that we have identified, or if it is a non-accurate (or accurate) output of an algorithm that we have yet to encounter/identify. If this is the case then we have undermined any possibility of making valuations about things.

    For exmple, if my calculator says "error" no matter what, I can't say it is broken, I just have to admit that I don't know what idealzed function it is running. The same objection can be raised if we appeal to telos in nature. Am I reaching for the apple because of the such and such a telos in me and in the apple, or am I supposed to be running a program that makes me drink gasoline and eat lead, and I have just malfunctioned? Since there is no way to tell by looking at the machines (i.e. the world) any appeal to telos had already been objected to on the grounds of inadequate knowledge.

    It seems like this paradox ultimately gives an argument against appealing to any sort of purpose or telos at all, becase we have no way of determining whether a thing is functioning according to A or malfunctioning according to B.

    Does that make sense?

    -Brian

    ReplyDelete
  44. @Crude I'd illustrate why, but holy hell - Ed, Grod and others ... ) have done so, repeatedly, now and over the course of months.

    Links please. I certainly don't remember your devastating illustration of my cluelessness, but perhaps I missed them the first time around.

    ReplyDelete
  45. gip,

    Links please. I certainly don't remember your devastating illustration of my cluelessness, but perhaps I missed them the first time around.

    Of course you don't, gip. You've been here, not getting the point of anyone's objections, for months now. It's not going to be solved by a link. The problem is, it's also not going to be solved by having the same conversation yet again. If a hypothetical creationist repeats, for the fiftieth time, "If man evolved from apes how come apes are still around?" and has resisted all attempts to have it explained to him, over the course of months, why his objection doesn't fly, the proper course isn't 'tell him yet again why it doesn't fly'.

    The proper course is to move on. Which is why I'm saying, expect people to move on at this point.

    ReplyDelete
  46. Crude, I'll believe in evolution the day I see a gorilla give birth to a human being. Until then, count me out.

    ReplyDelete
  47. I'm not sure why this is an argument against computationalism as such. I have a hard time seeing how any of this shows why it would be in principle in possible to program a computer to pass the Turing test, that is, to be behaviourally indistinguishable from a human being. And then the paradox being presented here would just be the same as Kripke's paradox applied to us humans.

    ReplyDelete
  48. "Hunt, if you really understand Dr. Feser's argument, restate it in your own words. Then we'll know whether or not you really do. So far, it appears obvious you don't."

    You mean I'm saying 125 but it's obvious the real answer is 5?

    Alright, I'll do this homework assignment later. Right now I'm out of time.

    ReplyDelete
  49. @Hunt:

    "Funny, that is what the clueless ones usually say. They say it to everyone who disagrees with them. They can't be wrong. The others just don't understand.

    Funny, that's what the really clueless ones say when I say it to them."

    The infinite back and forth is already implicit in my response. The circle cannot be broken. That you thought you had to spin it one more round just betrays your cluelessness.

    But the really irrefutable argument, is that you have ended your post with not one, but two smileys.

    Your smileys are powerless against our Trekkie Feser guns.

    ReplyDelete
  50. gip,

    See if you agree with this:

    1.) In the case of adding velocities, our actual operation is "quus" and not "plus", since, for velocities:

    x quus y = x,y if > c
    if not, = c

    2.) The question whether velocities are added or quused is not decided by looking at an adding machine. We can't just stare really hard at, say, an abacus and figure out Special Relativity.

    I take 1 and 2 as given, and I suspect you do too. But this means that an adding machine can't tell the difference between plus and quus, though the mind can tell that velocities combine by the latter process and not the former.

    ReplyDelete
  51. Anon @ 6:51 am,

    Sorry, your comment got stuck in the spam filter and I only just now saw it. Since it's now so far back in the thread, feel free to re-post it if you wish. I'll try to respond as soon as I get a chance. (Got to get an article done tonight.)

    ReplyDelete
  52. Hunt, rewording the argument is the best way to reply to anybody who thinks you're not getting an argument. I have found that useful whenever there is such an impasse. That way, you can say, "Have I gotten it right?". Once you nail it down, you can either agree with it or offer a cogent rebuttal, but after that, nobody will be able to say you don't get it. Like I said, it definitely appears you're not getting it.

    ReplyDelete
  53. Too funny: I told my wife just now about "quaddition," using the number 53 trillion decillion as the flipping point (past which the result is 5), on the ground I was sure she had never added numbers that large.

    She said, "Definitely, we're using quaddition. How else can we handle the national debt?"

    ReplyDelete
  54. David Brightly, let's suppose that an alien race, a superintelligent shade of the color blue, discovers your planet and starts investigating. First it comes across this "machine A" and through extensive testing determine that when you give it electrical input in a certain binary format it gives certain output of an "add" relationship. Then they discover a second "machine B", they do similar testing and discover when they give it like input, it uniformly has output of "0". They postulate that the meaning of A is "adding" and that the meaning of B is to return the additive identity value. When they tear B apart, they eventually discover that its internal structure is such that it ALWAYS will return 0, that's the way its structure is expressed.

    Along you come, and you look at the two machines, and you exclaim "no, no, you idiots, B is just a damn rock! It has no so-called meaning." What went wrong?

    When you have an adding machine that does "adding" with the right input, you want to suggest that "adding" is a meaning built into its structure. But the machine itself cannot tell you what the "right" input is. It cannot tell you "first press some of these gray buttons, then press the blue one with a +, then press more of the gray ones." Without those instructions implied, the machine isn't really an adding machine at all, it doesn't "do" addition. If someone uses it for a paperweight, it succeeds at that task too. He gave it a different input, and it had a different kind of output. If a third person uses it to brain his wife, THAT's what it "means" for him. Again, new input and different output. If a 4th person uses it to keep his 2-year old entertained because it's back is reflective, why it does that too, all without adding a thing. It means all of these, without once changing its clothes. Pretty versatile fellow for something whose additive meaning is intrinsic to its structure.

    The "meaning" you are finding presumes stuff independent from the machine itself.

    ReplyDelete
  55. CS is my field. One interesting thing about computers is that we can describe them, accurately, in some very different ways. A computer is

    a collection of silicon and plastic with electric current in it
    a collection of transistors
    a collection of gates
    a collection of bits, with a CPU that alters these bits in a pattern
    a collection of bytes
    something that runs assembler code
    something that runs an operating system
    something that runs C++ programs (or whatever language you use)
    something that runs your web browser, etc.
    something that helps you waste all your time on the Internet

    Thing is, each perspective is valid; many are useful or useless depending on what you're trying to do. And each is an _outside_ perspective. The computer doesn't tell you how to view it, how to interpret its activities, etc. You decide, and that's what makes it useful.

    ReplyDelete
  56. @Crude: so you can't actually point to any of these devastating retorts of yours. That is a bit suspicious, don't you think?

    I don't buy your analogy for an instant. A creationist making the argument you describe has failed to understand the mechanisms and nature of evolution. What corresponding thing in your universe have I failed to understand? If I am really so clueless it should not be difficult to produce an example.

    ReplyDelete
  57. GIP,

    so you can't actually point to any of these devastating retorts of yours. That is a bit suspicious, don't you think?

    I never made mention of 'devastating retorts'. I did make mention of various people over the course of months, and in this very thread, trying to explain something to you, explaining why your arguments fail, and you failing to get it. That you sit there and say "Well, I don't think they're right at all!" is to be expected, when what I'm maintaining is 'you've not gotten it, and odds are you're not going to, barring a change on your part'.

    PEBKAC, again.

    What corresponding thing in your universe have I failed to understand? If I am really so clueless it should not be difficult to produce an example.

    Producing the example for a clueless person is easy. Getting them to understand it? Not so much.

    Why, it's as if they're clueless or something.

    ReplyDelete
  58. She said, "Definitely, we're using quaddition. How else can we handle the national debt?"

    Damn, I used that joke myself in a public lecture recently. Now if I ever use it again, people will think I stole it from your wife! ;-)

    Ah well, great minds think alike, and all that...

    ReplyDelete
  59. GIP,

    The problem with the genetic example is the same problem that it has had when you've appealed to it in the past: If you are acknowledging that it amounts to an instance of irreducible "directedness" toward an end, you are implicitly committed to an Aristotelian position rather than a naturalist one. Hence you are opting for something like the "telos in Nature" approach referred to in the main post, which, as I have said, is not a defense of naturalism or materialism but the abandonment of naturalism and materialism.

    ReplyDelete
  60. Edward Feser said… I'm really amazed at how badly Vincent and GIP are missing the point.

    Torley, sure, but nobody's surprised that GIP didn't get it. It's the same point he has failed to see since he first showed up here. You can't miss it, because it's always accompanied by bragging about how his leet puter skillz enable him and him alone to see the mystical truth.

    GIP: Hi Hunt, yes, that seems to be the usual response. Disagreement can only be the result of failure to understand, or failure to read a couple thousand pages of Aquinas, or the like.

    Right, it's the usual response to ignoramuses. Look, the problem isn't that you disagree, it's that you clearly do not understand the problem. When people who are smarter than you or I say there's a problem, when theists and atheists alike say there's a problem, when naturalists and materialists say there's a problem, it's a pretty good bet there is a problem there. You and Hunt are like color-blind men turning yourselves blue in the face telling us there is no such thing as color. Well, that's pretty hard to convince those of who are staring at the colors, but it's worse than that. We're trying to explain color to you but we know you're not even in the ballpark because you keep retorting about sound!

    Oh, and the anti-intellectual pouting is right on schedule. Pretty hypocritical coming from someone who thinks his mad computer expertise makes him worth listening on anything with computers, but suggest that he get some expertise in philosophy to be able to discuss it intelligently and suddenly knowing what you're talking about doesn't matter any more.

    ReplyDelete
  61. I'm probably just pissing in the wind but here goes.

    If one's *basis* for grasping a formal operation F is just an empirical reckoning of all the instances of F he's seen or done in his life, then *for all he knows* he may as well have been doing an astoundingly similar operation F' which simply lay outside his data set. But we *know*, even on the basis of a single instance, that F (addition, modus ponens, defining succession, etc.) is simply not the kind of thing susceptible to "falsification" or "modification" based on empirical inquiry. If we don't know that about F, we don't know much of anything. Since we do know that about F, though, it's not based on physical analysis or physical parameters at all, which are always amenable to empirical analysis and redefinition in a way radically unlike F's completeness and perfection.

    To borrow from Aristotle, there is nothing at all physically different between AJ "the road from Athens to Jerusalem" and JA "the road from Jerusalem to Athens", yet there is a world of formal difference between them. As such our knowledge of *what* AJ and JA *are* is not in principle reducible to physical characteristics. This holds for all our knowledge of the appropriate kind. Our Popperian "third world" is a Leibniz Mill so vast and intricate that it admits of no complete walkthrough/analysis (since some here seem to think the point is that we can always just "crack open" the world like a laptop and see what's "really inside"). Nonetheless we have an undeniable grasp of what the Mill is doing, as it were, even and radically apart from its physical structure.

    Meanwhile watching the same typical objections come up about this argument provides a strange mix of depression and amusement. Leave it to our combox moguls to get the better of n00bs like Kripke and Ross, among others. Sigh.

    ReplyDelete
  62. Cogitator, I don't believe you're committing the biological reduction of potency to act with the wind here. If the likes of gip and Hunt won't listen, us newbies are learning a lot. Thank you.

    Bill

    ReplyDelete
  63. goddinpotty said… If I am really so clueless it should not be difficult to produce an example.

    Er, no, if you mildly misunderstood it, it would not be difficult to set you straight. Being SO clueless means it's probably impossible, because Ed has already given a very clear statement of the problem, plus followup (or is that "quus"), as well as comments and some really good examples from other posters. It's probably not possible to state the problem any more simply. Look, sometimes we just hit a blind spot with certain concepts. Nobody will be upset at your for not understanding, that wouldn't be fair. On the other hand, if you were to act like if you don't understand it, then everybody ELSE must be wrong…. well, even you should be able to grasp what's wrong with that.

    ReplyDelete
  64. Ok Dr. I think your last comment cleared things up a bit, for me at least.

    Does Kripke assume some kind of modernist/humean ontology regarding the natural world?

    That is, he assumes that cause and effect are not intrinsicly related to one another. No matter how much we analyse a machine M into its constituent parts M1,...Mn, its parts still behave like black boxes with intrinsicly unrelated inputs (cause) and outputs (effect). So, at the end of the day M will still be a (complex) black box, without intrinsicly related outputs and inputs?

    ReplyDelete
  65. An easier argument against computationalism is the simple observation that computers crunch numerals, not numbers. This is, after all, the core of Kripke's argument. The number one can be represented by '1', 'I', or an open logic gate. The latter is far more powerful than a scratch in the dirt because a properly wired collection of gates can represent many arithmetic theorems, though with some finite error rate. While 1+1=2 for all eternity, the binary machine operation which represents it, 1+1=10, has a finite error probability, unlike the theorem it represents. The confusion of numbers with numerals is as profound an error as computationalism itself.

    Also, Kripke's bizarre pseudo-arithmetic merely amplifies the difference between real numbers and their floating-point representations.

    For example, take 1,000 floating-point numbers of approximately the same magnitude, then order them and compute the differences between each pair. When you sum up the 999 differences you won't quite get the difference between the highest and lowest of the 1,000 numbers, while of course it's is always an exact equality with numbers. That's with double precision, while for quad it would take 100,000 numbers.

    When you've done lot's of numerical analysis programming you find many such examples of the numeral-number distinction.

    ReplyDelete
  66. @Feser: If you are acknowledging that it amounts to an instance of irreducible "directedness" toward an end...

    I don't. It is quite reducible to mechanism, which was the whole point.

    Hence you are opting for something like the "telos in Nature" approach referred to in the main post, which, as I have said, is not a defense of naturalism or materialism but the abandonment of naturalism and materialism.

    It's certainly not necessary to posit supernatural forces to make ribosomes work, so I fail to see how my example entails the abandonment of naturalism. I don't know exactly what you mean by "telos in Nature" but nature is full of purpose, all on its own. Which is the point.

    ReplyDelete
  67. GIP said: "I don't. It is quite reducible to mechanism, which was the whole point."

    Then, immediately after...

    "...so I fail to see how my example entails the abandonment of naturalism. I don't know exactly what you mean by "telos in Nature" but nature is full of purpose, all on its own. Which is the point."

    Forehead meets desk. You just can't make this stuff up.

    ReplyDelete
  68. @GIP

    ""You are trying to pull the same trick that grodrigues does (not that I think it's trickery on your part or his, I'm sure it is based on sincere belief -- but to me, it appears to be an attempt to sneak in the same invalid assumption through a slightly different door). Namely, setting up a disjunction between the "physical" and the "conventional".

    The best response to this is the example of genetic translation that I have offered up here before. The mapping from DNA triplets to amino acids is entirely "conventional" in the sense that it could easily be different. But it is also "physical", because the mapping is implemented by a population of tRNA molecules that bind to a segment of mRNA and a particular amino acid. Indeed, the translation conventions vary slightly in some organisms and in mitochondria.
    ""


    Besides what stated above by others, your example fails on another level as well.
    Sure one can say that genetic translation is 'conventional', but it is NOT truly so.

    Genetic translation depends DIRECLY on on physical properties of the system!!! UNDERSTANDING, does not.

    That is because DNA translation is a RECOGNITION process, not a process where something is understood. Yes there's a difference.

    The point it that the triples in DNA are NOT language with a 'meaning'. They are molecules that folloow their natural behavior.

    It is we who UNDERSTAND this as 'translation' and 'coding'... but in themselves they are just chemical reactions (although the *concept* of chemical reaction still is something that we understand).

    So the string og 'letters' in the DNA code that translates into a sequence of aminoacids really it is something VERY DIFFERENT than a phrase such as 'The cat is on the mat'


    In your example, yes the relation between "cat" and real cats is "conventional", but that convention is implemented by physically embodied brains that happen to share a language. Conventionality is not something separate from physicality.


    This does not explain the UNDERSTANDING of the words. It only explains the 'recognition of symbols'.

    A PC will RECOGNIZE the words 'The cat is on the mat' with a voice recognition program installed in it... it will NOT however, UNDERSTAND it.
    A machine does not understand anything AT ALL. A machine DOES. It is WE who *understand* the inputs and outputs.

    So it is TRUE that our recognition of symbols is a physical process.

    When you hear the word 'path integral', for example, your brain will recognize (if you know what it is at least) it by phisical processes and associate it with some 'memory' in your brain.

    Yet, understanding of what that exactly *means*, it's somewhat different.

    Again, a PC can compute an integral, it can do it much faster and better than I, as a matter of fact, but it will never understand what it means.

    ReplyDelete
  69. It's certainly not necessary to posit supernatural forces to make ribosomes work, so I fail to see how my example entails the abandonment of naturalism. I don't know exactly what you mean by "telos in Nature" but nature is full of purpose, all on its own. Which is the point.


    Too bad that was never Feser's argument.

    Thomism does not need God to directly operate the ribosome to make it work.

    Actually your example would even fail in a debate with an Intelligent Design supporter (Feser is not one of them, neither am I), since an ID-supporter would tell you 'God does not operate the ribosome, he just 'designed' it'.

    So much for that reply than!

    I don't know exactly what you mean by "telos in Nature" but nature is full of purpose, all on its own. Which is the point.


    Now now... be careful. You will make Mr. Dawkins and Mr. PZ Myers mad if you say abominations like "nature is full of purpose".

    Together with Mr. Dennett they work SO HARD in their books to say nature has no purpose whatsoever.


    Indeed your phrase is ironic if not comical.

    Telos in Nature entails exactly what YOU say. That Nature has a 'purpose' a 'directness' INTRINSIC to it.

    Mind you this 'purpose', Aristotle and Thomas also pointed out, is NOT in itself a conscious one (consciousness being something of a special case for higher animals).
    So when theists talk about purpose they do not mean that electrons make decisions, but that they have an intrinsic purpose in their nature… exactly as YOU stated.

    So basically you DO reject naturalism, even if you do not realize it yet :D

    Of course you can always join Mr. Dawkins and try to make away with that purpose. Well good luck with that then.

    ReplyDelete
  70. "Hunt, rewording the argument is the best way to reply to anybody who thinks you're not getting an argument. I have found that useful whenever there is such an impasse. That way, you can say, "Have I gotten it right?". Once you nail it down, you can either agree with it or offer a cogent rebuttal, but after that, nobody will be able to say you don't get it. Like I said, it definitely appears you're not getting it."

    Okay, I hope you will allow that I understand the initial rule-based paradox, which notes that any finite rule based system cannot distinguish between an infinite number of systems that can satisfy its rules. How this spoils inductive rule based systems like Peano axioms is not clear to me. It would seem that from them 'plus' is absolutely distinguished from 'quus,' but this is probably what Kripke meant in his confirmation-based communitarian skeptical solution to the paradox. Kripke only thinks there can be no symbolic meaning for an isolated individual. Right off the bat, the paradox MUST have a solution, whether Feser's or someone else's. Indeed, if there is no linguistic meaning, the paradox itself cannot be meaningfully expressed, but it is, and so there is a contradiction.
    Next, Feser applies the rule based paradox to machine processes. What intrinsic to a machine says that it's performing one function and not another. Earlier, but relevant at this point, Feser says:

    'The point for our purposes is that the “quus” example provides a useful illustration of how material processes can be indeterminate between different functions.'

    But it's not just a problem for a machine (or a mind, modeled as a machine, which is the point). Kripke has already conceded that there is no real meaning in any of the symbols or processes within an isolated individual (and certainly not a machine and double certainly not a molecule; this was the mRNA example). The question is whether anything can be salvaged about meaning in a community of interacting people, or machines, for that matter. Kripke (and I) think there is. The key is to abandon truth valuation and resort to verification and conflict avoidance or resolution. You may actually never entirely know that another person means plus or quus, but how much does that actually matter so long as you can verify meaning sufficiently to conduct transaction?

    If you heartily disagree, you might ask yourself why you think your interpretation is true and mine isn't. Well? Ironically, a lot of people actually think Kripke himself misinterpreted Wittgenstein when he created the paradox.

    ReplyDelete
  71. Mechanism, schmechanism. Naturalism, schmaturalism. Telos, schmelos. Purpose, schmurpose. Gip's has explicitly and repeatedly identified himself as a pragmatist: what works for him "in the real world" is what's true, for him, "of the real world." Hence, his––as I see it–– ad hoc terminological jerrymandering isn't very surprising. If one line of argument makes him a crypto-Aristotelian, he'll assert he's a hardcore materialist. If that, however, deprives him of real intentionality, as most thinkers on the issue realize it must, then he'll assert he's not that kind of extreme, marginal naturalist. If his phantom intentionality is shown again to be effectively Aristotelian, he'll claim such intentionality sounds too spooky or immaterial or "supernatural", whereupon he'll go back to hardcore materialism. Round and round it goes. Kind of like a dialectical halting problem, ironically enough.

    ReplyDelete
  72. Hunt: You may actually never entirely know that another person means plus or quus….

    So close, yet so far!

    The point is that if I myself just am a physical system (S(p)), I can "never entirely know" whether I myself mean plus or quus. But, as you note, since we obviously do know what we're doing with formal operations, it follows that we aren't just physical systems.

    Kripke's argument is a reductio against computationalism. His entire point is that rule-based algorithms (RBA) cannot in principle ground the kind of reasoning we can't deny is employed in everyday communication (C(r)). But since RBA are the essence of computational mechanisms (M(c)), M(c) cannot in principle be the basis for C(r). As such, M(c) qua species of S(p) are not the basis for our rational interactions and self-possession as competent language users.

    ReplyDelete
  73. Hi Codgitator,

    re: GIP

    Terminological gerrymandering or terminological confusion? I haven't engaged the fellow before, so I'll certainly defer to your conclusion. But I've never, never seen someone insist that they haven't "abandoned naturalism" and then, in the very next sentence, assert that purpose is an ingrained feature of the natural world. There's "pragmatism," and then there's just being thoroughly confused.

    ReplyDelete
  74. "The point is that if I myself just am a physical system (S(p)), I can "never entirely know" whether I myself mean plus or quus. But, as you note, since we obviously do know what we're doing with formal operations, it follows that we aren't just physical systems."

    I meant that in the context of Kripke's skeptical solution of community meaning, which is why I reference "another."

    I don't find the anti-computationalist argument seeking metaphor in modern machinery very convincing. A machine capable of supporting meaning as a human does will be vastly more complex than an algorithm instantiating plus or quus or any other function you can define exactly. It may very well contain meta-level knowledge about the functions it is performing just as you do. It will know it is performing one function and not another.

    ReplyDelete
  75. By the way, maybe Dr. Feser can say something about the requirement to prove non robot status. Quite apropos to this discussion. And extremely arduous. Or maybe there's some didactic point to it? I wonder if he's really plagued by robot attack that much.

    Well, poetic justice dictates that he may be.

    ReplyDelete
  76. Yes, Hunt, if we stipulate from the outset that i) we are vastly complex machines and that ii) a machine that knows what it is doing will know what it is doing, then your rebuttal is superbly apt and compelling. I just never knew establishing materialism was so easy!

    ReplyDelete
  77. Crude, Tony,

    Thanks for the replies.

    I think 'described as the binary representation of' can be explicated without resort to intentions as follows: We are clearly referring to a set, say 8 in number, of electrical potentials at various places in the device. If, in some fixed order they are LHHLHLLH, say, with L denoting a potential less than some fixed lower threshold and H denoting a potential higher than some fixed upper threshold, then we can describe them concisely with the natural number 105 whose binary representation is 01101001, ie, 1 corresponding with H and 0 with L.

    I'm not looking for a meaning or function or purpose for the device. I am merely offering a concise description of the way it behaves in the form 'if you set the inputs thus the output will be so and the inputs will be related to the output in a certain way determined by the structure of the device'. This certain way can be described in mathematical language.

    I think we all agree that we can't read off the purpose or intention of the device's designer from an inspection of its structure. But Ed, in the passage I quoted, seems to be claiming that we can't even say what the device does ('what the machine is doing' in Ed's words) without bringing in intentions, and it's this I take exception to. I offer my description as a counter example to this claim. It's early days yet, but no one has so far shown how this description involves the designer's intentions.

    ReplyDelete
  78. "Yes, Hunt, if we stipulate from the outset that i) we are vastly complex machines and that ii) a machine that knows what it is doing will know what it is doing, then your rebuttal is superbly apt and compelling. I just never knew establishing materialism was so easy!"

    Not a problem. That's what I'm here for! You only really need ii though, which seems pretty tautologous.

    ReplyDelete
  79. "The infinite back and forth is already implicit in my response. The circle cannot be broken. That you thought you had to spin it one more round just betrays your cluelessness."

    No, it's a recursion that bottoms out at any point when I decide to call you a silly twit (which I just did). Bye bye now.

    :-|
    >;-(
    :o

    ReplyDelete
  80. Hunt: "…which seems pretty tautologous."

    Bingo! #GettingIt

    ReplyDelete
  81. There is nothing in the physical properties of the machine that in any way references the meaning of what it is doing.

    All meaning is stripped out of an algorithm before it can be processed by a machine, an operation known as compilation. For example, the following two statements reduce to exactly the same algorithm within the memory of a computer

    (i) IF RoomLength * RoomWidth > CarpetArea THEN NeedMoreCarpet = TRUE

    (ii) IF Audience * TicketPrice > HireOfVenue THEN AvoidedBankruptcy = TRUE

    both, when compiled, will appear to the machine as an anonymous executable form such as
    IF a*b > c THEN d=1

    The computer will then perform the same internal operations whether its consequences are a visit to the carpet store or an embarrassing surplus in Max Bialystock’s bank account... http://rational-buddhism.blogspot.co.uk/2012/02/church-turing-deutsch-principle-and.html

    ReplyDelete
  82. It is you folks who seem to be entertaining private definitions of words. Since when does naturalism entail lack of purpose? Naturalism is the belief that nature is all there is, that there are no supernatural forces beneath or behind the natural world. Purpose (not some kind of overweening god-given purpose, but the kind of local purpose found in any living thing) is a feature of the natural world, so it should be obvious that there is no incompatibility between naturalism and purpose. Whether that constitutes "teleology" I will leave to you.

    ReplyDelete
  83. His Pragmaticness has spoken: "we people" have been judged casuists. "Ho gegrapha, gegrapha!"

    Now, if I had a dime for every time I've heard a naturalist tell me, "We make our own values, create our own purpose in a blind world"….

    ReplyDelete
  84. "I don't. It is quite reducible to mechanism, which was the whole point."

    "It's certainly not necessary to posit supernatural forces to make ribosomes work, so I fail to see how my example entails the abandonment of naturalism. I don't know exactly what you mean by "telos in Nature" but nature is full of purpose, all on its own. Which is the point."

    goddinpotty, these two remarks betray a very, very, very confused understanding of the philosophy of Aristotle and Aquinas. You really do not know what you are saying. Have you The Last Superstition? I think the book would help quite a lot.

    ReplyDelete
  85. gip:

    Here's an etymology of "purpose":

    late 13c., from O.Fr. porpos "aim, intention" (12c.), from porposer "to put forth," from por- "forth" (from L. pro- "forth") + O.Fr. poser "to put, place" (see pose). On purpose "by design" is attested from 1580s; earlier of purpose (early 15c.).

    As in: proposed design, propositional content, primal plan.

    You're telling me those notions are completely of a piece with big-tent naturalism?

    "Naturalism!"

    "You keep using that word. I do not think it means what you think it means."

    ReplyDelete
  86. Interestingly, "purpose" seems to have been coined in the very cauldron of Scholasticism!

    ReplyDelete
  87. Now, if I had a dime for every time I've heard a naturalist tell me, "We make our own values, create our own purpose in a blind world"….

    Humans are part of the natural world and so are their constructs. So a created purpose is still a purpose and still natural.

    ReplyDelete
  88. gip:

    Your creeping (or perhaps rank?) constructivism undermines the realism of the scientific consensus/worldview alleged to ground your protean naturalism. What we find in nature, according to you, is what we put there. But of course since we are just nature's tools, nature is writing her own purpose, tatooing herself with meaning, as it were. Cue pantheism and panpsychism, the scandalous secretly incestuous siblings of naturalism. You've said at lease twice now that purpose pervades nature, never mind where this puts you with respect to naturalism as a recognized ideology, yet now you're saying we qua entirely natural entities construct our own purpose. If we can, why can't everything else? Cf. incestuous metaphysics.

    ReplyDelete
  89. Potty writes

    "I don't know exactly what you mean by "telos in Nature" but nature is full of purpose, all on its own. Which is the point. "


    In order to explain what point, exactly, why don't you write up a little essay with that as your theme, explaining what you mean by the terms you are using, such as "purpose" and "nature" and what purpose implies in nature according to your definitions.


    Can you for example, define exactly what it is that you mean by the word "purpose" when you employ it with the possessive pronoun "its" in reference to "Nature"? Are agents necessary in your view for a, having or b, recognizing "purposes"? One of ? Both of? Neither?

    How do you distinguish what you call natural purpose from that which is natural and not part of any purpose?


    This blog and these recurring thread themes and disputes crossed my mind the other day as I was reflecting on the Medieval and Phenomenological notions of intentionality ... which seem at least to have the virtues being understandable. I mean you go into class, the professor draws an eye on the board and a tree in the distance and says " You are consciously intending the tree in a multilayered way as you fasten mentally on it as an object " ... etc.

    But what you mean by intention and even "purpose" is obscure to say the least.


    What exactly for example, is it that you think differentiates your view of "purpose" and intention from that of Feser and realists on the one hand and Rosenberg as eliminativist on the other?

    ReplyDelete
  90. "Humans are part of the natural world and so are their constructs. So a created purpose is still a purpose and still natural."

    In the immortal words of Voltaire "'Cause if we find we're in a bind we just make some shit up."

    note: not *that* Voltaire, but this one: Make s*** up.

    ReplyDelete
  91. @codgitator:

    Your creeping (or perhaps rank?) constructivism undermines the realism of the scientific consensus/worldview alleged to ground your protean naturalism.

    Constructivism is in no way incompatible with realism.

    What we find in nature, according to you, is what we put there.

    Well, most of nature is not constructed by us of course.

    But of course since we are just nature's tools, nature is writing her own purpose, tatooing herself with meaning, as it were.

    Nice metaphor, but misleading. "Nature" does not have a capital-P Purpose, it contains a vast multiplicity of purposes and purposive agents, often in conflict.

    I am not sure why this is so hard to grasp, seeing as it is so close to everyday common sense. Maybe that is the problem, I'm not using enough -ism words and similar pretentions. Eg, that the world is filled with purpose is common sensical, but turn it into "panpsychism" and it suddenly becomes a mystery and/or a controversy.

    ReplyDelete
  92. Yes, gip, that's your problem: you're not being pretentious enough. Sure.

    Slow your roll and try out DNW's idea.

    ReplyDelete
  93. Field test:

    Dear Naturalist, please check yes O or no X after the following claims.

    1. The purpose of the heart is to pump blood.

    2. The purpose of the eye is to see accurately.

    3. The purpose of the tongue is to speak truthfully and foster social harmony.

    4. The purpose of the genitalia is to produce offspring and foster social perpetuity.

    5. The purpose of the human organism is to know and love the Source of truth, goodness, and being.

    6. The purpose of these marks #%*//@! is to express that it's raining.

    ReplyDelete
  94. Dear Naturalist, please check yes O or no X after the following claims.

    What would be the purpose of that?

    ReplyDelete
  95. I see what you did there. :happyface:

    Even so, try the quiz. Humor me.

    ReplyDelete
  96. Nah, I don't really accept the terms (eg, that everything has a single purpose -- my genitalia, to take one of your examples, have at least four purposes -- reproduction, pleasure, urination, and impressing other men in the gym locker room. And let's not even get into the tongue...)

    ReplyDelete
  97. gip,

    Thomasts do not believe there is only one purpose for those things either. You can easily replace the purpose of with one of the purposes of.

    ReplyDelete
  98. Codgitator

    I'm no naturalist, but I'd have to respectfully disagree with 3. The primary purpose of the tongue is to taste food and to manipulate food for chewing. You could also say that phonetic articulation is a secondary purpose of the tongue. Speaking truthfully is not a purpose of the tongue, because it's an intentional act and tongues don't intend anything. People do.

    You could say that God gave us tongues to speak with, intending that we should use them to convey truth. But there's nothing in the tongue that's directed at truth, in the way that the eye is directed at the perception of color.

    ReplyDelete
  99. Hi Ed and grodrigues,

    I've been thinking about your comments. I want to make sure I've understood you correctly, so please bear with me. Let's say I land on a strange planet and discover what looks like a giant analogue computer. I discover what appears to be an AND gate and an XOR gate. Now I agree that for all we know, these gates might not have been designed by anyone; they might have arisen accidentally. But the AND gate is uniquely defined by four possible combinations of A and B (the inputs), so IF it was designed by anyone for some purpose, its logical function is unambiguous. Ditto for XOR.

    What about addition? Have a look at the 4-bit adder in this article: http://en.wikipedia.org/wiki/Adder_(electronics) . Let's say we find an analogue device on the planet with the same kind of arrangement. I will certainly grant that it doesn't exhaustively define addition - after all, it's only 4 bits. But there's no way that such an adder could be described as a quadder. It doesn't have anything in its architecture for producing funny output at higher values.

    You might object that an infinite number of mathematical functions are compatible with the inputs and outputs for the 4-bit adder. True. All that shows is that no finite material device can encode a formal mathematical operation which is valid over an infinite range (minus infinity to positive infinity). But adding is not defined by inputs and outputs. It's defined by a sequence of steps: first you do this, then you do that. And my point is that all the steps for adding are there in the flow of the sequential analogue device's operations. No extra steps are required, for the small range of numbers it can add. For any other mathematical operation, however, you would need extra steps (e.g. do something else for higher values in the case of quadding) or different steps. So the rational conclusion that a visitor to the alien planet would reach is that IF the device was designed for anything, it was designed for adding.

    Or am I still missing your point?

    ReplyDelete
  100. To say that genitalia have multiple purposes is to confuse the actual claim. Codgitator may have been too ambiguous, but it certainly is true to say that the purpose of the human reproductive system is to reproduce humans. In the same way, the excretory system is meant for excretion. Thinking that the argument turns on whether or not some body part can be used for multiple things, the tongue, for instance, is to misunderstand how purpose manifests itself in the case of living organisms. Biological science rests upon these crucial distinctions, and their very intelligibility rests upon philosophical concepts like finality.

    ReplyDelete
  101. E.R,

    Dr. Feser specifically addresses genitalia in TLS. It is agreed that the primary purpose of the reproductive organs is to procreate, but it is obvious that the male appendage is also used for urination. "Other directedness" is shown regardless the purpose one cites.

    ReplyDelete
  102. "All meaning is stripped out of an algorithm before it can be processed by a machine, an operation known as compilation."

    Unless the program is interpreted, in which case the original text is read and acted upon directly. I imagine that people here will still quickly dismiss this as just another meaningless process; however it is the start of a meta-level process that I mentioned above.

    "I you can't explain it simply, you don't understand it well enough."

    --Einstein

    ReplyDelete
  103. "Biological science rests upon these crucial distinctions, and their very intelligibility rests upon philosophical concepts like finality."

    At the molecular level many things have multiple function. Even more than that, many molecular machines change function through time. Behe's famous flagellum actually had origin in a cell secretory system. You may need to add a few epicycles to your theory to account for that.

    ReplyDelete
  104. At the molecular level many things have multiple function. Even more than that, many molecular machines change function through time. Behe's famous flagellum actually had origin in a cell secretory system. You may need to add a few epicycles to your theory to account for that.

    The very statement that something has a "function" is to attribute to it a final cause or range thereof.

    ReplyDelete
  105. Dr. Feser specifically addresses genitalia in TLS.

    Well, I sometimes talk to them too, but I don't feel the need to write books about it.

    ReplyDelete
  106. Is Feser or one you trolling us by pretending to be this "goddinpotty" fellow? He seems like the perfect caricature: ahistorical, confused, ideological, intransigent, sloppy, uncritical, and insolent. And it almost appears that he is actually proud of his philosophical ignorance and complacent as to his errors.

    What manner of creature is this? Surely, he cannot be real. It's like he is completely oblivious to centuries of philosophical discourse regarding causality and modern attempts to eliminate the necessity of resorting to "purpose" -- or how "purpose" is completely inconsistent with mechanism, a view universally held by actual philosophers.

    But then he tells us that he is a "pragmatist." Well then, perhaps he should write a paper explaining to all his fellow pragmatists in academia why they are wrong. Bring them all out of the analytic dark ages! Show them the light! Such brilliance must not be confined to a combox in the dark corners of the internet.

    Come on, gip, give it a go. Tell the world of philosophy why you are right and, quite literally, everyone else for the last 500 years has been wrong. Expound upon the wondrous insights that computer science and biology have furnished you. Such fields, alas, are beyond the reach of philosophers! Pray tell, oh gip, what is the hidden gnosis that you keep eluding to, oh prophet?

    ReplyDelete
  107. Daniel Dennett is considered a philosopher, is he not? Not a very popular one around here to be sure, but he has the professional qualifications. Here's what he has to say in Darwin's Dangerous Idea:

    The Darwinian Revolution is both a scientific and a philosophical revolution, and neither revolution could have occurred without the other.... In a single stroke, the idea of evolution by natural selection unifies the realm of life, meaning, and purpose with the realm of space and time, cause and effect, mechanism and physical law.

    Emphasis added.

    So Anonymous is simply wrong, as he could have verified himself with 15 seconds with Google. There have been many philosophers who see no inconsistency between mechanism and purpose, and his attempt to paint my very ordinary views as some kind of mystical gnosis is revealed as a pathetic lie. No wonder he refuses to use a name, even a pseudonym would be embarrassed by an effort like that.

    ReplyDelete
  108. Rogere Penrose (mathematical physicist) made a similar argument. But he employed "Godel's incompleteness theorems" to argue that the brain's/mind's functionality could not be completely reduced to an algorithm and thereofore was ultimately non-computable. His "quantum mind theory" is known as "orchestrated objective reduction."

    ReplyDelete
  109. Hunt is right; most biological parts end up having multiple functions, because repurposing is a primary mechanism of evolution.

    For a discussion of the problems of assigning fixed functions to biological structures, see here.

    ReplyDelete
  110. goddingpotty,

    > There have been many philosophers who see no inconsistency between mechanism and purpose, and his attempt to paint my very ordinary views as some kind of mystical gnosis is revealed as a pathetic lie.<

    I believe you're conflating "teleonomy" (apparent purpose) with "teleology" (real purpose). Neo-Darwinian evolution is considered by materialists to be a "telonomic" process, not a "teological one.

    ReplyDelete
  111. Mirabile dictu! The philosopher has spoken! Sophia has not so graced us since the forlorn days of Gemistus Plethon. Forsooth, it seemed that as if we were to return to the days of Plotinus, Porphyry, and Proclus.

    But then, oh horror everlasting, the prophets words -- they were false. This is not gnosis from the Demiurge. What chicanery is this?! Perhaps a mischievous henad at play? For he has failed us. Even Mr. Paisley, the initiate, sees the error: Dennent is talking about "purpose" in an equivocal sense than the way it is meant by the writers here -- telonomy and teleology are not the same thing. A verity that could have been verified "with 15 seconds with google." What a fool! What a pathetic lie! If only this false teacher would take a fortnight and contemplate the arcane texts of Introductory Philosophy. Then, perhaps -- if Apollo wills it -- he could, like Hypatia of old, come to gaze upon the Divine Intellect -- and stop embarrassing himself by his ignorance and complacency.

    ReplyDelete
  112. Wow, Am I in the Agora XD ???

    What gallant words sir, but I am afraid you will just make Potty even more pissed.

    ReplyDelete
  113. It's getting a little tweedy in here.

    ReplyDelete
  114. @Vincent Torley:

    "I discover what appears to be an AND gate and an XOR gate. Now I agree that for all we know, these gates might not have been designed by anyone; they might have arisen accidentally. But the AND gate is uniquely defined by four possible combinations of A and B (the inputs), so IF it was designed by anyone for some purpose, its logical function is unambiguous. Ditto for XOR."

    Let us grant then for the sake of argument that the inputs as well as the outputs are in binary form. You discover by inspection is that by feeding certain inputs x you get certain outputs f(x) and that the function x |-> f(x) happens to be what we know as the logical AND or XOR or whatever. You think "Wow, those aliens know elementary Boolean logic." Actually, you are already wrong here, but the example is blindingly simple, so let me move on.

    "What about addition? Have a look at the 4-bit adder in this article: http://en.wikipedia.org/wiki/Adder_(electronics) . Let's say we find an analogue device on the planet with the same kind of arrangement. I will certainly grant that it doesn't exhaustively define addition - after all, it's only 4 bits. But there's no way that such an adder could be described as a quadder. It doesn't have anything in its architecture for producing funny output at higher values."

    But of course it can. It implements *exactly* the same function as quus. Since we only have 4 bits and 2^4 = 16 just put

    a quus b = a + b if a, b <= 16,
    0 otherwise

    Want to add more bits? Up the cut off threshold. There you have it. The same machine implements addition and quaddition and a literally infinite number of other incompossible functions.

    "But adding is not defined by inputs and outputs. It's defined by a sequence of steps: first you do this, then you do that. And my point is that all the steps for adding are there in the flow of the sequential analogue device's operations."

    No, no, no, no, no. Even if we assume that the algorithm can be read off from the architecture, the same high school algorithm for computing addition for numbers up to 4 bits is *the same* algorithm that computes quaddition for numbers up to 4 bits, so you have not advanced one iota in refuting Kripke.

    "So the rational conclusion that a visitor to the alien planet would reach is that IF the device was designed for anything, it was designed for adding."

    That is the problem -- that conclusion is neither rational nor warranted, because on the basis of inspection of the machine alone, all you can say is that by a stroke of luck if you encode the inputs in binary format, you get back in binary form the addition of the two inputs. But you really do not know and cannot know, without appealing to the interests of the putative space aliens. Maybe they like to do quaddition. Maybe it implements some other bizarre alien function that by a pure stroke of luck, when the inputs are encoded in binary form, it has the same form and function as addition. There is just no way of knowing by looking at the machine alone. Sure, we can make inferential, probabilistic guesses, but that is neither here nor there.

    ReplyDelete
  115. Eduardo the gallant! How art thou? Let the fiend, gip of the flaccid gnosis, become pissed! His burlesque philosophy must be stripped bare and exposed as carrion. Justice demands it! Apollo must have his due. For gip, the lame heretic, has spoken of special insight received from Pragmatica in the fields of biology and computer science -- but has remained silent. What could this be but gnosis from the gods?

    Oh hunt, scourge upon reason, the tweed is necessary. This space must be purified of ignorance. How else do you expect us to receive Divine Wisdom? Your kind must be expunged, for your only hope is to be reborn again, but with adequate powers of reason, or theurgy. Yes . . . YES . . . that's it! Look to Iamblichus. He is the only hope for the deficient.

    ReplyDelete
  116. Technically speaking, real computers already do something very like quaddition. The limit is not 68, but 32,767 (at least if you use ordinary-sized integers). This is an aspect of the hardware. In this scheme, 32,767+1 yields -32,767.

    Is this a bug, or a feature? We tend to think it's a tolerable bug. But if you only look at the behavior, there's no way you would know that.

    ReplyDelete
  117. Someone please crack a window. The cigar smoke and book dust is killing me.

    ReplyDelete
  118. grodrigues said...

    "The same machine implements addition and quaddition and a literally infinite number of other incompossible functions."

    No, it does not. It implements a function with a finite domain and a finite range. Addition and quus are both extensions of that function. They are not identical to it. An computer adder does not, properly speaking, implement addition, but merely a function which is identical to the addition function on a finite subset of its domain.

    ReplyDelete
  119. True it appears that there is nothing intrinsic to the alien machine that says it's adding. Let's up the ante though.

    Say it's a machine capable of character recognition and it's programmed to read any number of digits off any number of pieces of paper twice and then add them.

    What I'm attempting here is not a closed argument, but rather a slow approach. Is it more difficult to say that this machine is not doing addition? And then continue the process, adding more layers of complexity. Say at the next level, when approached the machine recognizes any type of medium, grabs it of your hands, reads the numbers and adds them.
    If it becomes more and more difficult to say that the machine does not embody the function, is it warranted to suggest that an extraordinarily more complex machine might be said to embody its function as much as does a person while performing addition?

    ReplyDelete
  120. You know Anon you should pick a name XD it won't harm all that much.

    But poor GIP, all he wants is to... ... speak and be heard. u_u I will stop here, I had enough of heated arguments in the net.

    I admire your drive sir XD! I can't be like you or Dr Feser. I have no patience, no desire to forgive, no wish to repeat the arguments ad eternum XD.

    U_U yeah I am pretty damn gloomy. I need therapy.

    ReplyDelete
  121. Hunt said...
    "Say it's a machine capable of character recognition and it's programmed to read any number of digits off any number of pieces of paper twice and then add them."

    Thats a good idea. It is possible to construct a machine that implements addition. I mean the real addition function. It is no problem to construct a turing machine that just reads and writes from a tape by a finete set of rules. Then you implement in that turing machine the addition function. Now, given enough tape, you can compute any number as high as you want.

    What would Kripke say?

    ReplyDelete
  122. The interesting thing is that the difficulty in disproving that a machine embodies its function and the complexity of the machine are in direct proportion.

    Now, we have these amazingly complex machines (the point of contention) that also seem to embody their functions...

    Er, shall we add two and two?

    ReplyDelete
  123. "Say it's a machine capable of character recognition and it's programmed to read any number of digits off any number of pieces of paper twice and then add them."

    So, let's say it's a machine capable of doing exactly the thing people are disputing machines are capable of in the relevant says, and see where the argument goes? Great!

    ReplyDelete
  124. "Now, we have these amazingly complex machines (the point of contention)"

    L O L.

    The fact that you thing the point of contention here relates to complexity just demonstrates you don't even understand Kripke's argument. Even better, you do it by assuming and importing the very thing you're trying to argue for.

    You have a slow approach, with the slow aspect being of the "I can count to potato" variety. ;)

    ReplyDelete
  125. I dont the Kripkes argument again.

    It seems to me, that the problem, whether a certain machine implements some kind of function, is purely epistemological.

    Any being with enough epistemological resource can decide whether a certain machine implements a given function, just by looking at the inputs and outputs.

    E.G. a being that is eternal and cognizes past, present and future like in one moment can easily check all input-output correlations of a given machine, and then decide whether it implements that function.

    Whether one is able to decide such questions depends clearly on his epistemological resources.

    Granted that humans sometimes lack the necessary epistemic resources.

    Now, where is the bridge to ontology? Does anybody see it?

    ReplyDelete
  126. I don't know if the argument is necessarily ontological.

    I also think it is in a sense epistemological barrier. But perhaps the ontology could have something to do with the lack of Telos in the A-T sense in matter or in things.

    I am still afloat, but since I can't stop to think about it, because I am studying XD ....

    ReplyDelete
  127. @grodrigues "the same high school algorithm for computing addition for numbers up to 4 bits is *the same* algorithm that computes quaddition for numbers up to 4 bits"

    Surely this is incorrect. The algorithm as was described to me as a child made no exceptions for large numbers. The fact that the results for addition and quaddition are the same for numbers sufficiently small does not imply that the algorithms are the same. For in the case of quaddition I must at the last step check the size of the number and in addition I do not. The check is part of the algorithm.

    ReplyDelete
  128. @radp:

    "It is no problem to construct a turing machine that just reads and writes from a tape by a finete set of rules. Then you implement in that turing machine the addition function. Now, given enough tape, you can compute any number as high as you want.

    What would Kripke say?"

    The same thing. Even if we grant you a Turing machine with an infinite tape and you, its user, the omniscient power to know (somehow) that if you encode the inputs in a certain form, the machine will spit out their addition, the very fact that seeing that the machine does indeed implement addition depends on a *specific interpretation* of the inputs, shows that Kripke's argument stands unrefuted.

    By the way, I am just repeating what I have beenm saying all along.

    And lest some genius suggests that we feed our imaginary Turing machine with the outputs of a character recognition machine, the only thing I will say is that he would be just conceding Kripke's point by smuggling human intentionality via the back door with the two words "character recognition".

    ReplyDelete
  129. @reighley:

    "Surely this is incorrect. The algorithm as was described to me as a child made no exceptions for large numbers. The fact that the results for addition and quaddition are the same for numbers sufficiently small does not imply that the algorithms are the same. For in the case of quaddition I must at the last step check the size of the number and in addition I do not. The check is part of the algorithm."

    The algorithm implements both functions, since the integers were assumed to be <= 16, period. Do you have any experience with compiler optimization? One of the simplest optimizations is to throw away sections of code that the compiler can prove will never be reached. Draw the conclusions.

    Want to have more bits? Up the cut-off threshold in quus. Unbounded memory? See my previous post.

    ReplyDelete
  130. And the battle stalls as daylight recedes. It does not look good for our scrappy dilettantes, gip and hunt.

    Gip, the insolent oracle of sacred gnosis, could not distinguish the difference between telonomy and teleology -- and had not the rejoinder. While Hunt, his BFF, was caught smuggling intentionality into the argument by attributing the ability of character recognition to machines. What question-begging tomfoolery!

    Will they concede the point? Is this their end? Or will they decamp with a non-sequitur and a petulant dismissal, in order to live to fight another day? Stay tuned.

    ReplyDelete
  131. For gip, the lame heretic, has spoken of special insight received from Pragmatica in the fields of biology and computer science -- but has remained silent. What could this be but gnosis from the gods?

    Computer science and biology? Pray tell, what can this imply but Penrosian pattern recognition in relation to certain animute entities? Alas, forsooth and for goodness sakes, there be no gnosis from the gods, but a overlooking of quus.

    Penrose: I do not mean to suggest that all mathematical relations can be perceived directly as 'obvious' if they are visualized the right way--or merely that they can always be perceived in some other way that is immediate to our intuitions. Far from it. Some mathematical relations require long chains of reasoning before they can be perceived with certainty.

    But the object of mathematical proof is, in effect, to provide such chains of reasoning where each _step_ is indeed something that can be perceived as obvious. Consequently, the endpoint of the reasoning is something that must be accepted as _true_, even though it may not, in itself, be at all obvious.


    GIP: [silence]

    quGIP: Bosh. What others here have to say is neither immediate to my intuitions nor obvious, ergo it simply is not true. Besides, the quotation is irrelevant as it is philosophy that is under discussion, and not mathematics, right? Or am I deluded? Anyway, the purpose of _my_ points are, in effect, to state what is obvious to _me_, then provide a statement or two in order to enhance the lucidity of my point. That others fail to perceive directly or grasp my meaning is due to their eccentric, pretentious and nonpragmatic habit of kicking over the log and allowing themselves to be distracted by the incoherencies, contradictions and confusions swarming and squirming about. Concerned as they are with these things hidden from my light, they've grown accustomed to a darkness I cannot fathom. This explains why they're easily blinded by the brilliance of my points. Yawn.

    ReplyDelete
  132. 1. I didn't say "The only purpose", intentionally.

    2. Run my little survey by a hundred naturalist Darwinians and see how much unblinking affirmation of purpose you get.

    3. Gip: Dennett, the guy who expunges consciousness as an illusion (perceived by no one, to boot) and reduces intentionality to a pragmatic confabulation: yep, he's the guy I want in my corner about purpose and meaning. :s

    4. Hunt grants the physical indeterminacy of incompossible functions. Progress. (Besides, I like the Neoplatonic radio play. Heh)

    5. Gadp dodges the issue by smuggling in a prescient grasp of functions, ranges, common (universal operation, etc.). Petitio principii.

    ReplyDelete
  133. Codgitator,

    You know, it all makes sense when you think about it: the protean "pragmatism," his shifting definitions, the conflation of terms, etc. Gip must be a nominalist -- and I don't mean that he is an Ockhamist/conceptualist/termist -- I mean pureblooded nominalism. We're in Humpty Dumpty territory here:

    Alice and Humpty Dumpty are talking about the days of the year when one might get "un-birthday" presents. Humpty Dumpty speaks, looking at a page on which Alice has written a simple subtraction problem: 365-1 = 364:

    ". . . As I was saying, that seems to be done right--though I haven’t time to look it over thoroughly just now--and that shows that there are three hundred and sixty-four days when you might get un-birthday presents--"

    "Certainly," said Alice.

    "And only one for birthday presents, you know. There’s glory for you!"

    "I don’t know what you mean by ‘glory,’" Alice said.

    Humpty Dumpty smiled contemptuously. "Of course you don’t--till I tell you. I meant ‘there’s a nice knock-down argument for you!’ "

    "But ‘glory’ doesn’t mean ‘a nice knock-down argument’," Alice objected.

    "When I use a word," Humpty Dumpty said, in rather a scornful tone, "it means just what I choose it to mean--neither more nor less."

    "The question is," said Alice, "whether you can make words mean so many different things."

    "The question is," said Humpty Dumpty, "which is to be master--that’s all."

    ReplyDelete
  134. David,

    Isn't intention or purpose still included in setting the inputs, and even in the structure of the device to begin with?

    ReplyDelete
  135. Exactly, Anon, though you don't have to tell me. Not the first time I've felt like I'm in Wonderland in some of these comboxes. Smirking mouths making and remaking claims and insisting on not having been refuted or even answered, all the while floating disembodied from cogent terms and coherent syllogisms. Not that I'm a philosophical saint (maybe a ninja, though), and I know at least two of my recurring defects as an online interlocutor. Anyway, on we go. Down the rabbit hole.

    ReplyDelete
  136. Hunt: When two people discuss a bird in a tree, how do they know that their minds are actually representing the same thing? Or, let's say we consider the possibility of ever discussing anything with a robot. How will there be any congruence between our thoughts when the substrate hardware will be radically different? The answer is that we share a common external reality and the patterns our brains have learned correlate at an abstract level.

    The problem isn't how to know two minds are thinking the same thing. The problem is how to explain that one mind is thinking anything at all. A computer can be programmed to play chess, but from the computer's point of view — or rather, the computer's lack of point of view — it isn't "playing chess" at all, it's merely acting the way it has to act. We can say it's playing chess, but if you set two chess-computers playing against each other, it's simply a double mechanism where each side acts "compatibly" with the other. So to some extent (not far enough, but we can ignore that for now) human minds could act "congruently" without requiring natural telos. But regardless of compatibility or behaviour, I can really mean "addition" or "the bird is in the tree". I can hold specific concepts in my mind that cannot be "picked out" by mere congruence because "congruence" doesn't reach that far down.

    That's why complexity doesn't help. The argument is not, "This little machine is good for picking up pebbles, but it could never pick up big rocks", to which you can reply, "But a bigger, stronger version of the machine could lift rocks." The argument is, "This little machine cannot pick up anything, not even pebbles, so it could never pick up big rocks." A bigger, more complex version of a machine that can't lift anything is only going to be better at not lifting things. A non-teleology machine is not going to do teleology no matter how big you make it. It doesn't matter how well you can fake it from the outside. The human mind is not a complex "non-mind" that happens to be able to fake understanding to an arbitrary degree. The human mind can point, be directed at, addition itself, even if quaddition or some other fake would serve just as well in practice.

    You can adopt the eliminativist view and say that the human mind is just like a computer because it doesn't really mean anything anyway (just like a computer!). But if you actually mean anything, if human minds can hold concepts like "addition" or "the bird is in the tree", then you are acknowledging teleology.

    ReplyDelete
  137. Anonymous: He seems like the perfect caricature: ahistorical, confused, ideological, intransigent, sloppy, uncritical, and insolent. And it almost appears that he is actually proud of his philosophical ignorance and complacent as to his errors. What manner of creature is this?

    It sounds like you are describing the common or garden-variety teenager. Or garden-party teenager, I should say. It is (after all) a truth universally acknowledged, that a seventeen-year-old in possession of an Internet connection must be in want of an audience. Teenagers who know it all are not, as a general rule, bad sorts deep down; their youthful exuberance merely has little patience in waiting for us old-timers to catch up with their endless leaps of brilliance. If only we were as clever as they, there would be no problems. However, time is the great healer: in a few short years, he is bound to be astonished at how much the rest of us will have learned.

    ReplyDelete
  138. Just curious. What exactly do AI proponents expect a "robot WITH consciousness" to do that a "robot WITHOUT consciousness" cannot do?

    ReplyDelete
  139. grodrigues said...
    "....depends on a *specific interpretation* of the inputs, shows that Kripke's argument stands unrefuted."

    Ok. I can see that. But I thought that the interpretations of the inputs and outputs are given, because (to quote from the article):
    "The point for our purposes is that the “quus” example provides a useful illustration of how material processes can be indeterminate between different functions."

    The quus example seems not to work for sufficiently powerful intelligences.

    But its not supposed to be an epistemological problem, rather ontological.

    The argument makes sense, if you assume some kind of occassionalist ontology where there are no intrinsic powers or intrinsic cause-effect correlations, or "directedness" in nature. No natural teleology that is.

    But then, of course, if you state this aussumption explicitly, the result is trivial, because the whole world is just a big black box, for every kind of observer.

    ReplyDelete
  140. radp:

    I don't think things are as simple as you suggest, though it's heartening to see you grasp the import of the argument. For an even more focused form, do read James Ross's "Immaterial" essay at least once.

    1. It's not occasionalism, since it establishes at least one domain with undeniable real teleology, namely, human reason and thought. A great deals falls out from that conclusion, very little of it occasionalist.

    2. It's not simply an epistemological problem. Rather, it gets at the key demarcation between ontology and epistemology. For, in establishing human thought as determinate and formally 'exhaustive' in a way radically unlike 'the physical', it shows a real ontological cleft. It's not just the case that we don't know whether X is an adding or a quadding machine. It's rather the case that X functions as either (and according to countless other functions), and thus *is not* determinately *any one* of them at the exclusion of others. The formal indeterminacy of 'the physical' is a real result of its radical potency. Nature can run software (cf. Ross on Aristotle's Revenge) as well as we see it does because it is radically subject to depotentiation by formal perfection. Hylomorphism. This is radically unlike our acts of intellection, which are uniquely and undeniably determinate among incompossible functions. A tidal wave no more IS the pure sinusoidal wave function we use to describe it than the notes and paper of a symphony ARE the real music signified by it.

    ReplyDelete
  141. "The fact that you thing the point of contention here relates to complexity just demonstrates you don't even understand Kripke's argument. Even better, you do it by assuming and importing the very thing you're trying to argue for."

    The point of contention I mean is the proposition that minds arise computationally, i.e. are machine processes. The complexity part is just there to designate brains, which are complex machines.

    "While Hunt, his BFF, was caught smuggling intentionality into the argument by attributing the ability of character recognition to machines. What question-begging tomfoolery!"

    Nice, but that's also not what I meant. I meant character recognition as in the plain OCR software of your cheap printer/fax/scanner. Seriously, you didn't understand that? Then you didn't get the point of the entire exercise.

    ReplyDelete
  142. Codgitator,

    thx for your reply. I think I see your point.

    So, the argument runs:

    1. Matter is indeterminate in a certain respect.

    2. But Mind is determinate in the same respect.

    3. Therefore, Mind and Matter must be different.

    I still have a problem with premise 1.

    You say:
    "It's not just the case that we don't know whether X is an adding or a quadding machine. It's rather the case that X functions as either (and according to countless other functions), and thus *is not* determinately *any one* of them at the exclusion of others."

    I dont think this is true. Lets say X is a simple calculator with just one operation: +, and suppose someone tries every possible input on the calculator and sees what output he gets. If he writes it down, he will get a finite time series like this:

    T1: 1+1->2
    T2: 1+2->3
    ...
    TN: i+j->k

    The point is: if you assume a world in which a) cause and effect are intrinsicly correlated, and b) material objects have identity over time, then this series determines completly and exhaustively which function X implements. Namely a function which is identical to addition on a finite subset of the definition range of addition. This is different from addition and different from quus, because these functions have infinite domains (i.e. definition ranges).

    It is also completly besides the point, what the intentions of the engineers were. If they really indented to implement addition, they simply failed their task.

    An eternal being outside time could in principle determine every possible input-output correlation and therefore completly determine which function any given digital device implements.

    In a world with a) and b) in place it is merely a question of cognitive power whether it is possible to determine which function a given device implements.

    But if you drop a) or b) then the world is nothing other than a huge accumulation of individual facts without any general rule. In such a world no functions at all could be implemented.

    You could still salvage the conclusion, but premise 1 is not the way.

    ReplyDelete
  143. Radp, in a recent comment, highlights another remark of Ed's that I think needs careful interpretation. Ed says

    " The point for our purposes is that the “quus” example provides a useful illustration of how material processes can be indeterminate between different functions."

    For, on the face of it, the relation between the inputs and outputs of our putative 8-bit adder is quite determinate. We can write down its extension: (2,3)-->5, (11,7)-->18, etc, for all possible pairs of inputs. What can be indeterminate about that? The answer, I think, is that this relation itself can be given a compact description in more than one way. We can say it's the restriction of the plus function to the set 0..255 and we can also say it's the restriction of the quus function to the set 0..255, and plenty more besides. What we have is a bunch of different recipes (take the quus function and restrict it to 0..255, for example) for describing or constructing the relation which itself describes the doings of the circuit, at some level. So the indeterminacy here is in how we describe the goings on of some physical system, rather than the system itself, I think. And this shouldn't surprise us because we are familiar with the idea that a definite thing can fall under distinct descriptions.

    ReplyDelete
  144. 'The argument is, "This little machine cannot pick up anything, not even pebbles, so it could never pick up big rocks."'

    As far as I can see the argument is "this little machine cannot pick up anything, not even pebbles, and no other machine will ever be able to either," which is a good deal more pessimistic.


    'A bigger, more complex version of a machine that can't lift anything is only going to be better at not lifting things. A non-teleology machine is not going to do teleology no matter how big you make it. It doesn't matter how well you can fake it from the outside.'

    The entire thing turns on whether non-reducible teleology actually exists, which nobody seems to know (but many believe). To that extent, the above is question begging. It would be great (for your side) if it did exist. In fact it would be a trump card. But at the moment it's a little like me going to a horse race and saying to someone betting on another horse. "but your horse doesn't have my blessing, so it can't win. No matter what horse you bet on, it won't have my blessing, so it won't win." The disproof, of course, would be the horse winning, though it must be admitted that it might lose.

    'You can adopt the eliminativist view and say that the human mind is just like a computer because it doesn't really mean anything anyway (just like a computer!). But if you actually mean anything, if human minds can hold concepts like "addition" or "the bird is in the tree", then you are acknowledging teleology.'

    I'm acknowledging something, that I will admit. Whether it is teleology or some material process that we don't yet understand remains to be determined.

    ReplyDelete
  145. Here is an example that doesn't rely on restriction: suppose we have a device with one input and one output. The relation between input and output has the extension 0-->0, 1-->2, 2-->4, 3-->6, etc. Given the functions double(x)=x+x, and twice(x)=2*x, the device could be described as a doubler or a twicer. It could contain a doubling circuit or a twicing circuit. It could contain one of each and choose which to use at random, by detecting radioactive decays, say. Nevertheless, I contend that the relation between its input and output is wholly determinate.

    ReplyDelete
  146. radp:

    I'm glad we're at least not talking past each other!

    1. Why should I grant finite functions at all? The argument is about rule following, and rules are meant to apply to all possible cases. What you've effectively done with your TN series is carve out yet another incompossible function under the class of addition, quaddition, call it pladdition. That still means pladdition acts just like addition and quaddition, but after your ijk instance, reduces to quaddition and addition (to name only two alternatives, of course).

    2. Further, by alluding to time you recall a point I raised very early here, namely, that of Goodman's grue problem. Who's to say pladdition doesn't mean "like addition until ijk at spatiotemporal frame SP[ijk] but then like quaddition after SP[ijk]" or "like addition until the third iteration of ijk"?

    3. The KEY error here is to assume we grasp formal operations O(f) by induction. That's why you wave towards an infinite cognizer, as if an infinitely large abacus would allow us to induce the notion of addition after sufficient trials. We don't need to be infinite cognizers to grasp truly determinate formal operations O(f) (like addition, modus ponens, succession, etc.). It's simply wrong in principle to think we arrive at these O(f) by tallying up what happens after each trial. Since it is, however,

    A. the physical that we analyze in empirical study, and since it is

    B. not from empirical induction that we attain O(f), it follows that it is

    C. not by means of the physical that we know O(f).

    D. As such, the physical does not ground O(f), yet

    E. our minds undeniably instantiate O(f)'s in a way that excludes other O(f)'s *in the very definition*, and therefore

    F. our intellectual powers are not based on the physical (as the stuff of computational engines).

    You can't claim to "get" addition if you admit it might possibly admit of an exception in, say, a new series of trials or in "other universes" that follow different physical laws. We already do what your infinite computer does yet without running an infinite number of physical configurations that simulate the O(f) in question. THAT'S the crux of the argument: we get to the infinitely exhaustive, formally determinate nature of O(f) radically apart from the number and character of physical configurations involved.

    ReplyDelete
  147. David B:

    "…the indeterminacy here is in how we describe the … physical system, rather than the system itself…. we are familiar with the idea that a definite thing can fall under distinct descriptions."

    "…the device could be described as a doubler or a twicer."

    Much of what I just said to radp holds for my objections to you, as well. Please have a look, especially i) grueness and ii) the irrelevance of both physical configurations and iii) induction.

    I've cited you above to point out what I think you still keep missing, namely, that, by your own examples, IT IS NOT FROM THE PHYSICAL DEVICES THAT WE GRASP FUNCTIONS, BUT VICE VERSA.

    First of all, 1. if we are wholly physical beings B(p), it follows that our "descriptions" of things are also wholly physical B(p)*. Since you claim the physical A(p) is uncontroversially determinate, it can't be the case that our B(p)* AS B(p) are indeterminate, otherwise you ADMIT A(p) is indeterminate, AT LEAST in the case of B(p) and B(p)*

    Second, 2. how is that we know "a definite thing can fall under distinct descriptions"? If it is B(p) which ground the determinacy of O(f), then it is impossible for such determinate B(p) to generate incompossible O(f). (To head you off, perhaps, genus and species are not incompossible, so forget about man as "bipedal mammal" and "rational animal" etc.). In other words, if is A(p) which grounds O(f), then no series of examined B(p) can BOTH ground a particular O(f) AND ground an incompossible counter-function, yet that is exactly what you're saying happens.

    Finally, 3. why could the "twice-doubler" device D(td) be defined by either function? We would have to have the notions of twicing and doubling BEFORE we constructed D(td), yet you say D(td) itself adequately grounds the O(f) we want it to perform. This is incoherent, which machinephilosophy alluded to in his question to you. D(td) is, by your own admission, amenable to incompossible functions, yet we can only nestle them together in the logic gates if we first grasp how and why twicing differs from doubling *in the definition*. Every formal definition essentially abstracts from its physical instantiations, but you put the cart before the horse by cementing any O(f) to its particular physical parameters.

    An actor may marvelously instantiate the playwright's ideas, but that doesn't mean the actor is the source or explanation of those ideas. Likewise, just as computers qua B(p) can marvelously instantiate O(f), it doesn't follow that the latter derive from or are wholly grounded in the former.

    ReplyDelete
  148. But adding is not defined by inputs and outputs. It's defined by a sequence of steps: first you do this, then you do that.

    Vincent, it is not defined by specific inputs and outputs, but it IS defined in terms of the right sort of inputs: quantity. You cannot add "white" with "stuffy" and get any output with the algorithm. But since the inputs are (by definition) not part of the machine, the meaning "add" cannot be wholly given by the machine alone, it needs an interpreter outside the machine.

    ReplyDelete
  149. C(C),
    I'm afraid I don't understand your notation so I can't respond to much of what you say. A couple of points, however.

    I'm assuming we can use mathematical language, in particular the idea of a function, in describing the world. Doesn't Kripke's argument presume this? Indeed, it seems to be saying that we have too much choice in this. I'm not sure an investigation into how we grasp these ideas is relevant. Besides, it's likely to involve even more contentious issues. Likewise, is it relevant to ask how we know that a definite thing can fall under distinct descriptions? Surely we can take this as read for this discussion?

    Could I say, though, in reply to your third point, that I don't admit that the random twice-doubler is 'amenable to incompossible functions'. And this where we may be going wrong. I am identifying a function with an extension or graph. Different intensions (recipes, procedures, algorithms) such as doubling/twicing or plus/quus can have a single extension. You (and Kripke?) may be thinking of 'function' in this intensional sense. Does that sound right?

    ReplyDelete
  150. When two people discuss a bird in a tree, how do they know that their minds are actually representing the same thing?

    The same way the question itself arises about such representation. To know that two people are discussing a bird in a tree begs all the same questions, and assumes the same common representation that is supposedly in question.

    ReplyDelete
  151. @David Brightly:

    "I am identifying a function with an extension or graph. Different intensions (recipes, procedures, algorithms) such as doubling/twicing or plus/quus can have a single extension. You (and Kripke?) may be thinking of 'function' in this intensional sense. Does that sound right?"

    Even if we were to grant you that the intensional vs. extensional distinction is irrelevant, it still is the case that any machine implements an infinite number of incompossible functions and so it cannot be determinate in the way human thought is; the evidence is all over the OP and the combox discussion so I will not repeat it. To quote James Ross (heed Codgitator's advice to radp):

    "The same point again. I can reason in the form, modus ponens ("If p then q"; "p"; "therefore, q"). Reasoning by modus ponens requires that no incompossible form also be "realized" (in the same sense) by what I have done. Reasoning in that form is thinking in a way that is truth-preserving for all cases that realize the form. What is done cannot, therefore, be indeterminate among structures, some of which are not truth preserving. That is why valid reasoning cannot be only an approximation of the form, but must be of the form. Otherwise, it will as much fail to be truth-preserving for all relevant cases as it succeeds; and thus the whole point of validity will be lost. Thus, we already know that the evasion, "We do not really conjoin, add, or do modus ponens but only simulate them," cannot be correct."

    ReplyDelete
  152. grodrigues,

    Apologies if I am being obtuse but I don't understand what you mean by saying 'any machine implements an infinite number of incompossible functions'. Ed's post doesn't use this terminology at all. My original comment addressed his remark that 'you are not in principle going to be able to determine what the machine is doing in the absence...'. I took this to mean that it wasn't possible to describe what the machine does in the absence...' I think it is possible and gave an example. So either I'm wrong about my example or I'm not understanding what Ed means by 'determine what the machine is doing'. Where do you think I'm adrift?

    May I say that your quote from James Ross makes the point that human reasoning is determinate. I'm not disputing this.

    ReplyDelete
  153. Vincent:

    But there's no way that such an adder could be described as a quadder. It doesn't have anything in its architecture for producing funny output at higher values.

    It could be described as a quadder that doesn't work right instead of an adder. There's nothing about the physical facts that logically necessitate that it's one or the other. You might *infer* that it's unlikely that that's what its creators intended it to be, but take note: now you're sneaking in knowledge (or at least a reasonable guess) of its creator's intentions, which you gleaned from knowledge of your own intentions and by assuming that they are something like you, with your understanding of (and preference for) addition, your knowledge of implementing adders, etc.

    And my point is that all the steps for adding are there in the flow of the sequential analogue device's operations. No extra steps are required

    No extra steps? Well, let's see. You must assign the meaning of "1" to the relatively high voltage level, and "0" to the relatively low voltage level (or did they do it the other way around, or use something completely different?). Then you must take the four output "digits" and interpret them in the correct left-to-right order as a single combined output. Then you must interpret them as a base-2 number as opposed to, say, a string of 1s and 0s. Then you must do the same thing for your inputs (and, of course, you must give the adder the inputs in the correct way). Then, and only then, can you interpret the function as one of addition (or as poorly-implemented quaddition, or an infinite number of other possibilities).

    Of course, each time you use it, you're going to have different electrons in the registers than last time, so who's to say that the new electrons mean the same thing as the previous ones? And the whole thing only works if you pass the "right" sort of current through it, but there's nothing physical about the thing that implies its for adding and not for emitting smoke when connected to a 10,000 volt battery. Again, you might infer that the creators of it think something like you and are rational, and that that probably wasn't their intention, but again, you're now bringing in knowledge of intentions.

    ReplyDelete
  154. Hi Ed:

    Suppose we say instead that there is what Buechner calls a “telos in Nature” that determines that the brain really is following this program rather than that -- the program for addition, say, rather than quaddition? In that case we would have some end or purpose intrinsic to the natural world that determines which program the brain instantiates, which would eliminate the occasionalist problem the appeal to God as programmer raised.

    Unless the "telos in Nature" is intrinsic to the human person specifically, as opposed to only Nature external to ourselves, don't we end up with another occasionalist problem, only with Nature as programmer (even if a genuine programmer, with intrinsic and irreducible telos that can assign meaning) rather than God? Doesn't that still (incoherently) imply that our thoughts don't have intrinsic meaning?

    ReplyDelete
  155. grodrigues,

    To help me out in this, could you explain the sense in which a bicycle implements an infinite number of incompossible functions?

    ReplyDelete
  156. @David Brightly:

    "Apologies if I am being obtuse but I don't understand what you mean by saying 'any machine implements an infinite number of incompossible functions'. Ed's post doesn't use this terminology at all. My original comment addressed his remark that 'you are not in principle going to be able to determine what the machine is doing in the absence...'. I took this to mean that it wasn't possible to describe what the machine does in the absence...' I think it is possible and gave an example. So either I'm wrong about my example or I'm not understanding what Ed means by 'determine what the machine is doing'. Where do you think I'm adrift?"

    Let us take your imaginary machine. By inspection, we know that it takes one input x -- for example, reads it off from a sequential tape of 0's and 1's as holes and non-holes and outputs f(x) in the same form, e.g. by punching holes in a sequential tape. Once again, assuming we have enough time and that the machine has no unbounded memory (a pretty reasonable assumption) you, can by inspection, draw the complete graph x -> f(x) of what the machine does. Now the question is: what does the machine compute? You cannot answer *this* question, *unless* you know what the inputs are supposed to mean: they can be sequences of bits, the most straightforward interpretation given my somewhat slanted description above, they can be unsigned integers in the usual binary format, or signed integers in two's complement representation, or floating point numbers in the IEEE 754 standard, or sequences of characters in some specific encoding (ascii, utf16, any of the variants of iso 8859, Hong-Kong hkscs, etc.), or machine code for some assembler language, etc. The variations are endless. And that is why I said, that the machine implements an infinite number of incompossible functions, and thus it *cannot* be determinate in the way human thought is: here I am alluding not to the post itself but to James Ross influential article "Immaterial aspects of thought". Kripke's argument expanded in the OP is a variant establishing this indeterminate-ness of computers: one cannot by inspection of the machine alone, establish that it is doing addition or quaddition or any number of other infinite possible variations.

    ReplyDelete
  157. grodrigues,

    Thank you for the challenge. In its most general form the machine maps a sequence of bits onto a sequence of bits. If by inspecting its innards we find it consumes the input 8 bits at a time and outputs 8 bits and that each output is independent of all inputs but the last (maybe it resets itself every 8 bits) then we can describe what it does as a function from 0..255 to 0..255. This accurately predicts what pattern of eight holes is punched out for every eight read in. And this more refined description entails the description in terms of bit streams. Either description is true. And neither requires interpreting the inputs and outputs as anything other than what they can quite reasonably be said to be, viz, certain patterns of holes in a tape. There is no need to know what the designer had in mind. We just figure out what it does from an account of its structure and our knowledge of physics. It is, after all, a deterministic system, and presumably sufficiently stable to give consistent results. Else it can't be said to compute a function at all, I think. Having done this we may spot certain patterns in the extension of the function we have derived. Oh, look, it's squaring, we might say, or even quusing. But that's just another, perhaps more concise, way of saying the same thing as quoting the graph.

    ReplyDelete
  158. Is it reasonable to conclude that Potty is evading the question as to what, exactly, it is that he means by natural or nature's "purpose", and how he determines that "nature" has "purposes"?

    It seems that if there are going to be these serial challenges to Feser's notions concerning the meaning of purpose and the implications he draws from that meaning, the challenger should at least make it clear what it is that he means *when he uses the same terms* but asserts a different conclusion as following

    This is beginning to look as though Codgitator is right, and there is some frantic fence sitting going on here: as a certain commenter realizes that once the doctrine of purpose he espouses is viewed within the environment in which he places it, it all either 1. gets a little Hindoo-ish ultimately as distinctions between nature and artifact are also jettisoned ... or 2. plants him unwittingly, or at least unwillingly, at Feser's conceptual feet.

    ReplyDelete
  159. "what does the machine compute? You cannot answer *this* question, *unless* you know what the inputs are supposed to mean ..."


    Your references to ASCII and EIA code prompts me to consider that an alien has 90 yards of ASCII punch tape commands and no machine tools, nor servos, and has never seen a helicopter blade machined to template form.

    "Knowing physics", what's the meaning?


    And though it's not dispositive of anything in particular it reminds me that we know humans, and still can't read Etruscan ...

    ReplyDelete
  160. Deuce says (of an adder with inputs greater than 57)
    "It could be described as a quadder that doesn't work right instead of an adder."
    Surely this is absurd. It's like describing a bicycle as a bus that didn't come out of the factory right. If, however you go on helpfully to say in what way the putative quadder doesn't work right, you end up saying that where it doesn't quad correctly it adds. So it adds on all inputs. You've described it as an adder.

    He asks,
    "who's to say that the new electrons mean the same thing as the previous ones?"
    Neither I nor Vincent, I suspect, wishes to say that they mean anything. We don't understand why other commenters seek meaning in a distribution of potential. Associating the natural number 105 which has binary representation 01101001 with the pattern of high and low potentials LHHLHLLH, is merely a convenient descriptive act.

    ReplyDelete
  161. David B:

    You seem to be, as you note, hung up on the extension issue. I can't help feeling however that you're simply not adequately versed in the philosophical issues involved to get this argument. That's actually a compliment to the effect that you won't let a good argument get in the way of good engineering. You seem to embody a certain strain of objection to this argument which I've seen crop up time and again in my grappling with it over the past few years.

    Your objection, I believe, amounts to this: "Saying a physical system is indeterminate, is just absurd, since the system is obviously doing something; we can even graph it!" This strain of objection seems to be rooted in an instinctual fear or mockery of the idea that the physical is formally indeterminate AS IF that claim meant physical reality is just an illusion or a sheer "black box" (à la radp). While such a reflex is understandable, it profoundly misses the point.

    So, to make a long story short, are you even aware of Duhem and Quine's arguments about "points on a curve" and the underdetermination of evidence? Do you realize how little a graph establishes against this argument, indeed, how well graphing plays into its hands? If not, no wonder you're stuck on an ineffectual reference to extension vs. intension.

    Likewise are you aware of the import Wittgenstein's rabbit-duck illusion has for physicalism? In the following letters, what about THEIR PHYSICAL FEATURES IN AND OF THEMSELVES makes the letters semantically determinate? "GODISNOWHERE"

    ReplyDelete
  162. Surely this is absurd. It's like describing a bicycle as a bus that didn't come out of the factory right.

    It's absurd only in that it makes no sense that any rational person would intend it for that purpose. Using only physics, you can't distinguish between an adder or a badly designed quadder, or a bicycle and a bus that didn't come out right. It's only because of your knowledge of rational agents and how they think that you're able to infer what it's probably intended for, and that various alternatives are nonsensical.

    ReplyDelete
  163. Codgitator,

    I think I put forward a cogent argument in my first comment that Ed's
    assertion that you can't describe what a machine is doing without reference to the intentions of its designer is false. I gave a counter example. Nobody has yet rebutted my counterexample. Psychologising one's opponent is not an argument. But thanks for the compliment!

    Regarding 'graph', apologies, I fear may have misled readers by using the word in a technical sense to mean just the set of (argument, value) pairs of a function. The graph of the function that maps the people in my street to their houses can't be plotted as a curve on paper, though it's pretty determinate, I think.

    Is underdetermination of evidence relevant here? It's not that we are formulating a theory about our adder. We are just trying to summarise the evidence of our eyes and instruments in a compact way. Don't we just see that a certain output goes with a certain input? We aren't trying to theorise why.

    I doubt the semantic underdetermination of symbols is a relevant issue either. Nowhere in my adder example do I talk about what the patterns of electrical potential might mean. The whole tenor of my comments is to avoid interpretation and rely on description. Now you could say that this is missing the point. But no one has, yet.

    Could someone show me what I have misunderstood about Ed's assertion or how my counterexample is misconceived? And in what way is thinking about this question in extensional terms inappropriate? What does it miss or get wrong?

    ReplyDelete
  164. Deuce,

    Let's not pursue the absurdity issue. Do you see my point about convenient descriptions versus meanings?

    ReplyDelete
  165. Do you see my point about convenient descriptions versus meanings?

    Yes, but it simply helps to further illustrate the point that Ed made in the first place. "Convenience" is quite obviously not a physical property that exists in the absence of rational agents with intentions. You're just bringing in more knowledge of rational agents and their intentions to infer what something was meant for, specifically the knowledge that they're likely to assign meanings that are convenient to them (which you reasonably assume to be similar to what would be convenient for you).

    ReplyDelete
  166. >> You're just bringing in more knowledge of rational agents and their intentions to infer what something was meant for <<

    No, that's not right. I make no inferences as to meanings. I merely find an interesting mathematical relationship between a pattern of electrical potentials in one place and a pattern of same in another place. I'm just a physicist doing an experiment, recording the results, and stumbling on a pattern. The convenience is merely mine---I encode patterns of input and output potentials as pairs of natural numbers, and then notice a pattern within that set of pairs.

    ReplyDelete
  167. Ἀμμώνιος ΣακκᾶςMay 11, 2012 at 3:42 PM

    Alas, gip has flown the field, leaving Hunt alone to battle the onslaught. He had not an answer to the "devastating critique" of his amateurish equivocation between different meanings of "purpose": telonomy, which is apparent purposefulness, and teleology, which is intentional purposefulness. When challenged, he remarked that his position is "common sense" and that this is how people actually live their lives. SAPIENCE! And we can concur: most people, like gip, aren't philosophically astute and confuse or conflate various concepts all the time.

    With this response, matters weren't looking good for gip. Facing reincarnation or daunting theurgical studies, he despaired. But then -- as if as a gift from the Divine Intellect -- he had an answer: his view is "pragmatic" -- a finer philosophical word if there ever was one! Now his words and concepts are permitted to shift meaning and content when convenient -- and though such absurdity has not been matched since the days of Ζήνων ὁ Ἐλεάτης and his silly paradoxes, gip believed that he had secured victory. With this, he departed from the comment thread, not to be seen again until such burlesque is once again in need! After all, he has truly slain the beast of Aristotle and not some bowdlerized mythical variant. O frabjous day! Callooh! Callay! He chortled in his joy.

    Gip, you had asked earlier for an example of a devastating response to your lunacy. Turning coat here and failing to respond to criticism uncovering the category errors hidden behind your juvenile derision and brash countenance is a good example.

    ReplyDelete
  168. Believe it or not I actually have a job and a life that sometimes is more pressing than engaging in philosophical arguments, especially when those have degenerated to the kind of feces-flinging on display here.

    So, let's see -- you introduce out of nowhere a distinction between "teleonomy" and "teleology" and then giving me grief for not adhering to this all-important divide. Well, sorry to disappoint, but I have no obligation to adopt your or anybody else's vocabulary, especially when it not-very-subtley encodes the very topic under discussion. Sorry, I'm not trapped quite that easily.

    Let's say there is some kind of "real" purpose that is distinct from "apparent" purpose. Presumably we have some way to detect the former and distinguish it from the latter, or it would be an utterly useless concept, and you wouldn't want to trying to force an utterly useless concept on the world, would you? So pray tell how does one go about identifying real purpose as opposed to the ersatz kind?

    In fact, the distinction is meaningless. There is purposive behavior. Like pornography, it may be hard to define but we know it when we see it. That's because we have [[http://cogweb.ucla.edu/Discourse/Narrative/michotte-demo.swf][specialized cognitive mechanisms]] for doing so. In fact those mechanisms are exploited by one kind of ersatz purpose, the kind displayed by fictional characters in the movies. But that is presumably not the distinction you are trying to make, because that's no different from the difference between an apple a picture of an apple.

    Gip, you had asked earlier for an example of a devastating response to your lunacy. Turning coat here and failing to respond to criticism uncovering the category errors hidden behind your juvenile derision and brash countenance is a good example.

    First off, you don't seem to know what "turning coat" means. I suppose you must mean "turning tail". Becoming a turncoat means defecting from one side to the opposition, and I don't think you think I've joined up with your side, do you?

    Even substituting the correct cliche, your statement makes no sense. My failure to respond is a good example of what? A devastating response? How is a lack of response simultaneously a response? And why am I responding or failing to respond to myself?

    Or are you simply unfamiliar with the basics of English because you spend so much time on shiny ten-dollar words like "teleonomy"? That's the glint of fool's gold that so attracts you.

    ReplyDelete
  169. goddinpotty (May 11, 2012 6:28 PM)--My failure to respond is a good example of what?

    goddinpotty (06/21/2009 12:20:00 PM)--"Answer the question or I'll take it as a tacit admission that you can't think your way out of a wet paper bag."

    ReplyDelete
  170. Touche!
















    ----------------------------------
    Between you people and me... this talk is so high on logic of lack of it, I am totally lost.

    ReplyDelete
  171. Ἀμμώνιος ΣακκᾶςMay 11, 2012 at 10:55 PM

    Gip, my acolyte, don't be so deceptive! The henads will not look upon it so kindly. Zeus may send Apollo to disembowel you for your obstinacy and effrontery. And don't fabricate such stories. You have been told for months that you are using "purpose" in a sense equivocal to the way in which everyone else here uses it. Sure, the actual signifiers "teleology" and "teleonomy" were not used until recently, but what they signify has been discussed for some time, only under different names.

    This is precisely Codgitator's complaint. You employ concepts in a way that belie clear distinction, in effect shifting the meaning of your claims with every claim. Instead of taking this distinction seriously, which philosophers of all stripes use, including your precious Dennent, you brushed off every attempt at correction. This is why you sound absolutely loony and clueless when you cite Dennent as a philosopher who believes in purpose in the way that we are using it.

    So I might have made an error in using the wrong word, but you've been an ingrate with a bloated ego and unwilling to learn anything. Seriously, if you haven't figured out what teleology is after being here for months, you aren't here for serious dialogue -- you're just a troll.

    Now get to those theurgy lessons before I sic an Archon on you!

    ReplyDelete
  172. gip:

    By pointing out your "pragmatic" M.O. as an interlocutor, I am NOT trying to impugn your character. I'm just saying it's very bad form, weak sauce, to boot.

    Also, if you think the distinction between teleonomy and teleology is ad hoc, it belies a truly ill- or underinformed grasp of the issues on your part. (Alastair was very on-point to raise the distinction.) If you also think the distinction is spurious and self-defeating, welcome back to the Aristotelian homestead! My impression is that you are basically contemptuous of the pretensions of philosophy, as you see them, compared to the real knowledge had by cogsci. As such, you don't philosophy very seriously. Get refuted? Bat it away as supernaturalism. Get corrected? Bat it away as philosophical adhockery. Get informed by your opponent with a key insight? Bat it away as common sense. See the protean-pragmatic-Wonderland M.O. many of us are complaining about? IlAlas, I'll wager you "don't" and shame on me for my "ad hominem". Sigh. (O_.)

    Again: slow your roll and take DNW's advice above.

    ReplyDelete
  173. Dear Naturalist, please check yes O or no X after the following claims.

    1. The purpose of the heart is to pump blood.

    2. The purpose of the eye is to see accurately.

    3. The purpose of the tongue is to speak truthfully and foster social harmony.

    4. The purpose of the genitalia is to produce offspring and foster social perpetuity.

    5. The purpose of the human organism is to know and love the Source of truth, goodness, and being.

    6. The purpose of these marks #%*//@! is to express that it's raining.


    None of these are literally true.

    ReplyDelete
  174. None of these are literally true.

    What a world.

    ReplyDelete
  175. If there is no such thing as irreducible teleology then what you call teleology might as well be teleonomy anyway; it's simply a matter of reducing it. If that is the case then everything is apparent purpose only. Some might find this repellent, which is of course irrelevant to its truth value. Finding it repellent is probably its own form of equivocation, since purpose and meaning in this case isn't the meaning life has for you that allows you to get out of bed in the morning. You can still be motivated to get out of bed with only teleonomy.

    ReplyDelete
  176. @Codgitator (Cadgertator) said...

    "1. Why should I grant finite functions at all? "

    I think you should look up, how a "function" is defined. The definition range is essential to the function definition.

    Maybe it is irrelevant in most practical applications of mathematics, but it is already essential for some key theorems in funcional analysis, set theory etc... were the proofs depend crucialy on the notion, that functions with different definition ranges are different.

    A calculator with a finite input has a finite definition range, and is therefore different from addition, quus, or any other function with an infinite definition range. And if some minimal assumptions about the natural world are true, then this device determines one and only one function.

    If you want to deny, that there are finite functions, then please define your non-standard use of the word "function".

    "3. The KEY error here is to assume we grasp formal operations O(f) by induction."

    I have never said that.


    BTW, if someone hasnt already noticed, I dont buy the whole epistemology/world view behind the argument. I am just interested in the merits of that argument, as a philosophical argument from a different world view. And it seems to me, that they are not much. A key premise is refutable while staying in the same world view.

    ReplyDelete
  177. This comment has been removed by the author.

    ReplyDelete
  178. @radp:

    "I think you should look up, how a "function" is defined. The definition range is essential to the function definition."

    Essential? Yes, a function is commonly defined as a triple (A, B, R) where A and B are sets (domain and range) and R a relation, a subset of the cartesian product AxB satisfying a couple of conditions. But it is common mathematical practice to either extend or limit the range of the function, as usually the context makes clear what specific function we are talking about.

    But this piece of pedantry completely misses the point. Codgitator is talking about rule following which typically have an infinite number of instantiations (e.g. axiom schemas, deductive rules, etc.). To translate his point 1. in your pedantically precise language, since a machine has finite inputs and finite outputs, it implements a function with finite domain and finite range with an infinite number of incompossible extensions: addition, quaddition, pladdition, whatever, and thus it is indeterminate, contrary to human thinking that must be determinate. When we add, we do addition, not quaddition or pladdition. When we apply modus ponens, a deductive rule with an infinite number of instantiations, it *must* be modus ponens, for applying modus ponens is truth-preserving for all cases that instantiate the rule. Applying modus ponens cannot therefore be indeterminate among other possible rules, some of which are not truth preserving. But rule following, or an arithmetical operation such as addition, arithmetics and syntax being closely related, *is* indeterminate if it is machine implemented (this is the point of Kripke's argument), therefore human-thinking cannot be machine-implemented.

    ReplyDelete
  179. @grodrigues

    "it implements a function with finite domain and finite range with an infinite number of incompossible extensions: addition, quaddition, pladdition, whatever, and thus it is indeterminate"

    This applies not only to a machine but to the finite function itself. Every finite function has an "infinite number of incompossible extensions". Therefore, according to *your* reasoning, every finite function is indeterminate. This results not from any material property but from the function itself.

    See, "pedantry" (rigorous definitions that is) is sometimes useful after all.

    ReplyDelete
  180. radp:

    I suppose "grant" was the wrong word, but I think you missed my point when I noted that any finite function you define just carves out Kripkean space within addition, i.e. gives yet another variant on his quus point. I meant, why should I grant the range is at all relevant, since even such functions are susceptible to Kripke's objection.

    You need not have 'said' what wrote about your position and induction, but what have said amounts to the target of my critique. You are claiming that if we just "look at what the machine does", we can know what it does. But what you and David B keep missing is that even when we graph the extension of your or his favored finite function, the extension is still susceptible to an infinite number of incompossible descriptions. Have you actually read Ross's essay or Kripke's book on Wittgenstein? Do you understand the Duhem-Quine problem of underdetermination and "points on a curve", much less Goodman's grue problem?

    The worldview you reject is in fact computational mechanism, since that view claims functions just are given by finitary heuristics (rules). You know that's wrong and thus effectively grant Kripke's point. Have you read Quine's *From A Logical Point of View*? For it is that book that he scorns realism as fiction on par with Greek polytheism and considers all physical objects merely conventional constructs. That's the worldview you reject; as such Kripke is your ally. You just don't like how he gets there.

    To date, though, the objection I've heard over and over from you, David B, hunt, gip, et al. to the claim that "the physical is formally indeterminate" (the premise 1 you wrote for me) is effectively: "Nuh-uh!"

    ReplyDelete
  181. Codg:

    "But what you and David B keep missing is that even when we graph the extension of your or his favored finite function, the extension is still susceptible to an infinite number of incompossible descriptions."

    I dont miss that. grodrigues mentioned something similiar. My answer to him applies here too. To have an infinite number of "incompossible" extensions is a property of the function, and that property makes nothing indeterminate, since every function, not only finite functions, but every function would be indeterminate then.

    ReplyDelete
  182. @ goddinpotty,

    > Let's say there is some kind of "real" purpose that is distinct from "apparent" purpose. Presumably we have some way to detect the former and distinguish it from the latter, or it would be an utterly useless concept, and you wouldn't want to trying to force an utterly useless concept on the world, would you? So pray tell how does one go about identifying real purpose as opposed to the ersatz kind? <

    If you make no distinction between the two, then you are forced to conclude that evolution is a purposeful and goal-directed process. (Of course, if you do make a distinction between the two, then you are forced to conclude that your personal behavior is striclty a teleonomic process (i.e. it's a process that only has the appearance of purpose).

    "A teleonomic process, such as evolution, produces complex products without the benefit of a guiding foresight."

    (source: Wikipedia: "Teleonomy")

    ReplyDelete
  183. No, radp, the functions would not be indeterminate in and of themselves and *in the definition*; that's why all pure functions are mutually incompossible. Nonetheless, any number of "data points" observed and said to fit some function would be indeterminate among incompossible functions, since, again, it's not *from* data points, graphs, trends, OR ANY PHYSICAL CONFIGURATIONS WHATSOEVER that we derive or grasp pure functions. I'll grant that you're not missing the point we're saying you're missing, but you're still missing the point.

    And I reiterate the questions I posed in bold above. Answering them would be a sign of good faith, and might help us stay on the same page. Cf. also my points above about the road to Jerusalem, GODISNOWHERE, actors & playwrights, etc.

    ReplyDelete
  184. If you make no distinction between the two, then you are forced to conclude that evolution is a purposeful and goal-directed process.

    You people seem to be prisoners of your words, as if they were the masters and you were the servants rather than the other way around.

    IOW, what makes you so sure that "evolutions is purposeful" is a well-posed question with a definite answer one way or the other?

    The reality is pretty clear: evolution is an undirected process that produces purposeful creatures as emergent products. That is pretty wondrous and strange, but not that hard to understand. Whether you call it teleology or teleonomics or teleobollocks is a matter of very little interest. If those words help you understand the reality, than that's great; if they hinder understanding, they should be abandonded.

    ReplyDelete
  185. His Pragmatics, the master of language has spoken! If our use of words don't help him engage reality (not to mention don't help him be king of every combox), he's free, like a Nietzschean pimp, to change their meaning as he sees fit. Fortunately, his own prescription means we can do the same with his words. So never mind what gip says, he only means for us what enriches our mastery of reality. WAR IS PEACE! 2+2=5! Not even pure functions are determinate and true! Huzzah!

    ReplyDelete
  186. @radp:

    "This applies not only to a machine but to the finite function itself. Every finite function has an "infinite number of incompossible extensions". Therefore, according to *your* reasoning, every finite function is indeterminate. This results not from any material property but from the function itself."

    Yes, there are an infinite number of extensions of any function whatsoever, but the argument does not hinge on *that* mathematical fact and what it establishes is only accidentally connected with *that* mathematical fact.

    ReplyDelete
  187. @Codg:

    "Nonetheless, any number of "data points" observed and said to fit some function would be indeterminate among incompossible functions, since, again, it's not *from* data points, graphs, trends, OR ANY PHYSICAL CONFIGURATIONS WHATSOEVER that we derive or grasp pure functions."

    I think it is irrelevant to the point in question, from where we grasp pure functions. The question is: Can a physical device said to implement one and only one function, or is it indeterminate in this respect?

    I say: If you take some minimal assumptions about the physical world for granted, then it is in principle even *knowable* (therefore also determined) which function a given device implements.

    Take the calculator example: You know *every* possible input-output correlation. The function which this divice implements is therefore also determined.

    Of course this is different, if you assume a continous process from which you take some measurements. From these alone you cannot say, which function this process represents. But this is just because of a lack of epistemic power. It says nothing about ontology. Just because you cannot know doesnt mean, it is not determined.

    Lastly, I think this whole issue rests on a ill-defined notion of function. Every well-defined predicate should be such, that it divides the world in two: in those things of which it can be affirmed, and in those things of which it can be denied. If now someone comes and says: it is not determined whether this thing is K, where K is supposed to be a predicate, then this simply proofs that his predicate K is ill-defined.

    "Have you actually read Ross's essay or Kripke's book on Wittgenstein? Do you understand the Duhem-Quine problem of underdetermination and "points on a curve", much less Goodman's grue problem? "

    No, I havent dealt with these books or problems. Neither did the blog article mentioned them as prerequisites to understand the argument.

    ReplyDelete
  188. One last remark

    I said: "...Just because you cannot know doesnt mean, it is not determined."

    Coming to think of it. This confusion between ontology and epistemology is not uncommon with Kripke. Like "Imaginable means possible" and such things.

    ReplyDelete
  189. @radp:

    "I say: If you take some minimal assumptions about the physical world for granted, then it is in principle even *knowable* (therefore also determined) which function a given device implements.

    Take the calculator example: You know *every* possible input-output correlation. The function which this divice implements is therefore also determined."

    It was already explained why determined in the sense you are employing here cannot do the work of undermining the argument.

    ReplyDelete
  190. goddingpotty,

    > The reality is pretty clear: evolution is an undirected process that produces purposeful creatures as emergent products. That is pretty wondrous and strange, but not that hard to understand. Whether you call it teleology or teleonomics or teleobollocks is a matter of very little interest. If those words help you understand the reality, than that's great; if they hinder understanding, they should be abandonded. <

    Evolution is the product of mechanism and chance blindly playing themselves out. And since you are nothing more than mechanism and chance blindly playing themselves out, then any purpose you believe you have must necessarily be deemed illusory.

    The problem here is that you are presupposing free will - a presuppostion that your worldview precludes. On the materialist worldview, there is no free will because there is no final causation. As such, all your intentions and purposeful behavior must necessarily be deemed illusory.

    ReplyDelete
  191. radp:

    [I type a lot on my mobile phone which accounts for erratic misspellings and elisions you might notice. For some reason I also can't seem to copy text in these comboxes on my phone, hence I sometimes must paraphrase, as I shall in a moment. It also makes it very tedious for me to interact on on threads over 200 comments at this blog, so I better get this in while it's easy! If I'm more scarce here soon you'll know why, even apart from my weekend ending soonish.]

    You mention the machine for which we know every possible input-output argument but fail once again to see why this is irrelevant to answering Kripke's argument. To recall your ijk pladdition machine, it is precisely the finitude of its range which makes it impossible to know whether it's pladding, quadding, adding, or whatever SINCE ALL THOSE FUNCTIONS PRODUCE IDENTICAL INPUT-OUTPUT RESULTS, *until* we reach an argument where they diverge. It may just be the case that your machine's computational range never reveals their eventual but intrinsic divergence. "I can see what it's doing in every possible case!" you protest. "Yes, but within *that* range of arguments it's doing exactly what an adder an a quadder do," Kripke answers. Just replace your ijk parameters with those Kripke proposed for quus, or vice versa, and I hope you'll see why.

    And again, this is not merely an epistemic problem, since, ontologically, the same physical system can BE in states identically referable to functions which mutually exclude each other *in their definitions* (BE formally indeterminate among 'competing' formal descriptions), but our minds cannot BE in such a state of indeterminacy with respect to our grasp of the truth-saving conditions of any pure functions. Knowing the entirety of the functions inputs and outputs is irrelevant (viz. have you or could you done every possible instance of addition or modus ponens?); as such, it is you yourself that are confusing an ontological with an epistemological issue.

    Feser doesn't mention every possible nuance and reminder in any post, which is why his comboxes are enriching fill-in-the-details spaces. Thank you for admitting you're not well versed in the authors/issues I asked about. Have a look, it's an exciting area of research!

    ReplyDelete
  192. Codgitator,

    So, if I managed to program a computer to be behaviourally indistinguishable from a human being, you would conclude that it cannot possibly have a mind because of what you know about computers?

    ReplyDelete
  193. @Codg

    Thx for your reply Codgitator. But I think we have come to an impasse regarding the question whether material devices can determinately implement a function.

    OT: I am just curious about your philosophical background. Is your world-view captured by some known "-ism"? Aristotelian-Thomism, perhaps?

    ReplyDelete
    Replies
    1. radp:

      I concur, but there are worse things in life, and it's certainly not the first time I've reached an impasse on these issues.

      I'll show you mine if you show ne yours: I'm a pretty convinced Thomistotelian but of a decidedly semiotic bent. John Deely, Walker Percy. I'm also heavily influenced by Stanley Jaki, and, as a result, Duhem and Pascal (and Montaigne). Donald Keefe's _Covenantal Theology_ was transformative for me, though he makes no bones about the incomplete conversion of Thomism as received, so I'm self-conscious of trying to see if I'm a "die-hard" Thomist or not. Along the same lines, from Keefe and Ross (and the Smithy) I've gained huge respect for the whole Scotistic tradition, so there's that as well. I will say, though, that Ross's writings, and his "Immaterial Aspects" essay in particular, made me want to "be a philosopher". So while I don't consider it a pet argument of mine, I recognize ny own very high esteem for it, and strive to be as well read, cogent, and distinct as Ross was. In a word, let's just say I'm a Catholic cadger/codger who cogitates at depth.

      Delete
  194. Ἀμμώνιος ΣακκᾶςMay 12, 2012 at 10:27 AM

    You people seem to be prisoners of your words, as if they were the masters and you were the servants rather than the other way around.

    "When I use a word," Humpty Dumpty said, in rather a scornful tone, "it means just what I choose it to mean--neither more nor less."

    "The question is," said Alice, "whether you can make words mean so many different things."

    "The question is," said Humpty Dumpty, "which is to be master--that’s all."

    ReplyDelete
  195. Rupert:

    I would say that any putative understanding on its part was not based on its physical structure, but that would just be to reiterate Kripke's, Ross's, Feser's (and Aristotle's) and my own point wrt other cognizers. Intellection is determinate in a way radically unlike physical computation, so I'd have to conclude that a machine you tell me is wholly physical does not reason as we do. That it does what I do in every case is interesting, but goes nowhere in addressing the argument at hand. After all, do you think your calculator is conscious of adding, much less of finishing its math homework? Or that your ruler is conscious of measuring, much less of building a house? Complexity is a typical red herring fallacy here.

    Indeed, your mention of "behavior" plays right into Quine's hands (cf. gavagai problem), which, in turn, ties into Ross's argument, so you're barking up the wrong tree from the get-go. We don't need simulators to *be* just like us to help us get tasks done (i.e. we don't need actors to be playwrights in order to have a good show, as I noted above).

    Further, consciousness vis-à-vis machines is a cloudy excessive hypothesis (Ockham's razor and all that). It makes an otherwise manageable and profitable research project geeky, ideological, and far too costly. I have no *reason* to assume this machine is more conscious than that machine, but I'd be mad to assume this or that person is less conscious than I am. A strange but performatively undeniable (cf. retortion) asymmetry in the human experience.

    Cf. Alastair's very astute question about strong AI and consciousness, above. Read some Mortimer Adler and Herbert Dreyfus on intellection and computers, too.

    ReplyDelete
  196. What exactly is your solution to Kripgenstein's paradox in the case of a human cognizer?

    ReplyDelete
  197. Ἀμμώνιος ΣακκᾶςMay 12, 2012 at 11:02 AM

    Gipparoo,

    The distinction is not a set of words that we are prisoner to. Did you not read Codgitator? We don't like it. Naturalists are the ones that made it up so that they can have their mechanism and their "purpose" too. It is to make sense of YOUR view of the world by providing an explanation as to why purpose appears in a world where mechanism precludes it. This is precisely the way that Dennet understands it -- and that's why you looked like a fool when you brought him into the conversation. To repeat: this has been basic knowledge to philosophers and a huge problem for mechanism for the last 500 years. To dismiss it without consideration makes you deeply unserious.

    ReplyDelete