Joseph
Bessette on criminal sentencing
laws and retributive justice, at Public
Discourse.
The Catholic Thing on the
late, great Michael Uhlmann. Requiescat in pace, Mike.
At The New Atlantis, Benjamin Liebeskind reviews Neo-Aristotelian
Perspectives on Contemporary Science.
At The Spectator, Roger Scruton
looks back with gratitude at an annus
horribilis.
Jez Rowden’s
Steely
Dan: Every Album, Every Song will be released next month. Ultimate
Classic Rock on the
great Eagles/Steely Dan cross-reference.
What Is It Like to Be a Philosopher? interviews
Scott Soames. Soames’s new book The
World Philosophy Made has just been released.
Sabine
Hossenfelder of BackReaction on the
crisis in physics, mathematical
laws of nature, and cognitive
bias in science.
Richard
Marshall interviews
philosopher Thomas Pink, at 3:16.
Glenda Satne
reviews
Alex Rosenberg’s How History Gets Things
Wrong, at Notre Dame
Philosophical Reviews. Bookforum on flummery
in the name of neuroscience.
Sexual
immorality and the decline of civilizations: Kirk Durston on J. D.
Unwin’s Sex and Culture.
Forbes on the
final season of The Man in the High
Castle. The Wrap explains
that enigmatic ending.
J. K.
Rowling, Ricky Gervais, and the right to say the obvious. Theodore Dalrymple comments at City Journal, and Kyle Smith provides
an update at National Review. Spiked
wonders whether
the tide is starting to turn against trans lunacy.
Regulating
pornography: Pascal-Emmanuel Gobry on why
it should be done, at American
Greatness. Terry Schilling on how
it can be done, at First
Things. Gerard Bradley on the same theme, at
Public Discourse.
Capturing Christianity interviews philosopher Tim
Hsiao on the topic of sexual morality.
Massimo Pigliucci reviews
Paul Feyerabend’s Philosophy of Nature,
at Philosophy Now.
At Harper’s, Lionel Shriver on
left-wing linguistic legerdemain.
Los Angeles Review of Books on
David Bromwich on the left and free speech.
Church
Times reports on Rupert
Shortt’s new book Outgrowing
Dawkins: God for Grown-Ups. Ben Sixsmith at Arc Digital performs an autopsy
on the New Atheism. Prospect on agnostic
Stephen Asma on our need for religion.
Also at 3:16, Richard Marshall interviews
philosopher of biology John Dupré.
At Science, James Gunn on Isaac
Asimov at 100. Asimov’s Foundation
series is
being adapted for television by Apple.
When
science fiction gets things wrong, at The New Republic.
Times Literary Supplement reviews
John Gribbin’s and Lee Smolin’s new books on quantum mechanics. IAI TV
interviews
philosopher of physics Tim Maudlin about the same subject.
There is a
website devoted to the work of Australian Thomist Austin Woodbury.
At Church Life Journal, Shaun Blanchard on
the smear word “Jansenist,” and Richard Yoder on
a pope who was resisted.
At Notre Dame Philosophical Reviews, Darrell Rowbottom reviews K. Brad
Wray’s Resisting Scientific Realism.
Fr. Mitch
Pacwa interviews
philosopher Jennifer Frey on EWTN
Live.
James
Franklin on
Aquinas’s realism for The Sydney
Realist.
I guess this counts as an open thread of sorts?
ReplyDeleteAnyway, I have a question about Feser's argument for the immateriality of thought or formal thinking processes. It might be a confusion on my part, but here it is:
The claim is we cannot know, just from physical facts alone, whether we are adding or quadding. And likewise nothing about the physical facts of a calculator/computer could tell us whether it was performing addition or quaddition. All we would get would be a series of inputs and outputs which are perfectly indeterminate between addition and quaddition, since both would lead to exactly the same outputs (except if someone entered the arbitrary quaddition number).
The claim is that even if we looked at the program we would not be able to know whether the computer was performing addition or quaddition. This is one problem I have. I can see how mere inputs and outputs, standard behavior, cannot show whether it is addition or quaddition that is being performed. However, by looking at the program we are able to see exactly what function the computer is performing. If the program does not specify anything about one of the operands being 57 or greater, or X (whatever makes quaddition different from addition, that is), then the computer is obviously not performing quaddition, but rather addition.
An adding function and a quadding function would give us the same results for every number lower than 57, but they would nevertheless be different. If the program does not specify anything about 57, then it is not ever carrying out quaddition since it is not really considering whether the number is lower than 57; it is not even attempting to follow the quadding function, and is in fact only performing the adding function. We can inspect a program to see whether the computer is carrying out addition, or if it is doing quaddition instead, so we can indeed know what determinate operation the computer is doing simply by looking at its program and checking what the function is. Because quaddition is functionally the same as addition except for operands equal to and higher than 57, a quadding function would have to specify exceptions or functions for numbers equal to or higher than 57, and if it doesn't do that then it is simply adding.
This could work as a functionalist response. Noam Chomsky might have already made that point before - his suggestion, as I remember it, is that we can know whether it's addition or quaddition by analyzing the causal process that gives the result. While the result might be the same for both operations (except when the number 57 or higher is involved), the operations and causal process are obviously different. So there can be a determinate fact of the matter of whether it's addition or quaddition, simply based on the physical function that carries out the operation.
Sure, we can tell what the computer is doing. We just can’t tell what it’s supposed to be doing. Maybe the software was supposed to be adding, but the programmer was incompetent and messed it up. Or maybe the data isn’t a program at all — maybe all those bytes were a JPEG that just coincidentally acted like a quadding program when you executed the file instead of opening it in Photoshop. Or maybe cosmic rays flipped some bits in your RAM and it’s random data that coincidentally acted like a quadding program. Or something else.
ReplyDeleteYou see, you say we can read the code to see what the program does (or is supposed to do), but that’s cheating — a language (programming or otherwise) is intentional, so if you are treating something as writing in some language you have already crossed from mere physical facts to intelligent interpreting. Mere physics just tells you that you have some magnetic pattern or flow of electrons, not whether the pattern means image data, or program data, or a fluke coincidence, or anything else.
I agree with Mr. Green here. You are assuming, Atno, that we can look at a computer program and tell that it is or is not checking whether a sum is greater than 57 or not. This is to assume that what the variables represent are numbers, that one of its subroutines (if it 'is' quadding) is adding, etc. The argument is intended to deny even all that, however.
DeleteThe appeal to causal processes does not help, in any case, since if you look at the causal properties of the calculator, there is a determinate fact about the matter, and that fact is not consistent with the hypothesis that it is adding. For sufficiently large values, the calculator does not return 'sums'.
Atno has a point. The Rossian claim is that the machine computes some function but that it's indeterminate which. A prerequisite of this claim is that we have decided what physically constitutes an input and an output and how numbers are physically represented by machine states. But then a sufficiently deep physical analysis of the machine will identify the function. This is independent of the intentions of the machine's maker, its causal history, and so on, none of which is relevant.
DeleteBut that’s the point of Ross’s argument — any talk of inputs, outputs, functions, etc. all has to presuppose some interpretation. You can’t get that from the physics alone; you have suppose that context pre-physics. If our minds were (only) physical, then we couldn’t get off the ground. Since we can — we are, obviously, able to get that interpretative context from somewhere — our minds then cannot be merely material.
Delete@Mr. Green
DeleteI think what he's trying to say is that the physical units of the machine such as number units and rotors and displays etc, which help make the calculation go on physically, are necessarily related to the proper interpretation in such a way that it's impossible to have a machine without a determinate function because that function is expressed through the physical determinate wiring.
Except they aren’t necessarily related — the indeterminacy at the heart of the argument means that the same function can be implemented with different physical setups, or different functions can be implemented using the same physical setup. The relation between matter and meaning is contingent. The physics is determinate insofar as whatever happens physically is what happens, but the meaning isn’t. Rotors and circuits and low voltage and high voltage do not describe a functum, what is performed — they describe what is happening. Adding more physics (e.g. saying it’s the operation of the machine + the operation of your brain) is just more of what’s happening, not what any of it means.
DeleteWe can surely agree on the presuppositions as to inputs and outputs and representations without regard to any theory of mind. We can't begin to discuss whether the machine implements plus rather than quus without this in place. Ross then runs an argument inspired by Kripke intended to show that the machine is indeterminate between plus and quus as to the relation between inputs and output. But there obviously has to be something wrong with Ross's argument---it's hard to say quite what because the argument itself is unclear---because, as you say, the physics is determinate.
DeleteDavid Brightly: We can surely agree on the presuppositions as to inputs and outputs and representations
DeleteWell, we surely can — because we have minds. The machine can’t.
We can't begin to discuss whether the machine implements plus rather than quus without this in place.
Precisely. You agree with Ross that we need to bring in such presuppositions; the physics alone does not suffice.
it's hard to say quite what because the argument itself is unclear
That is the argument. You can be, in your mental action, determinate in a way that we all agree mere physics cannot. Therefore your mental faculties cannot be (merely) physical.
(I don’t know if you’ve seen Feser’s expanded treatment of Ross’s argument. You might also be interested in Revisiting Ross on the immateriality of thought and other links in that post.)
@ David Brightly
DeleteThe Rossian claim is that the machine computes some function but that it's indeterminate which.
He is actually at pains to deny this (whether you mean that it is epistemically or ontologically indeterminate which function the machine computes). The machine does not compute any function.
A prerequisite of this claim is that we have decided what physically constitutes an input and an output and how numbers are physically represented by machine states. But then a sufficiently deep physical analysis of the machine will identify the function.
If in our analysis of the machine we advert to how we decided to interpret its physical state, then we are not doing a physical analysis of the machine.
My target is the argument Ross gives in 'Immaterial Aspects of Thought', section II, 'The indeterminacy of the physical', page 140 here. Ross applies an argument of Kripke concerning human thought to physical systems. It's ironic that Kripke's original argument was intended to cast doubt on the reliability of reference of terms like 'plus'. See for example here. So Ross has borrowed an argument that casts doubt on the other half of his thesis, that mind is determinate! In so far as K's argument succeeds it's because in thought we have only access to historical outcomes. We have no access to any underlying medium of thought which we might have independent reason for thinking reliable. Such as matter.
DeleteWhat's ironic? One philosopher comments on another, taking one aspect of his book to be valuable and correct, and another aspect to be flawed. Happens all the time. In fact most readers of Kripke find his sceptical solution inadequate (and not Wittgenstein's) and prefer preserving the idea that humans can actually add etc., though many of them (rightly) think his exposition of the paradox is brilliant.
DeleteThis is all why Ross wrote Section III of his paper.
I am puzzled by the "ironic" claim, too, since Ross is very explicit (in that very paper) that he thinks Kripke has mislocated the skeptical implications of the plus/quus case.
DeleteIf the machine computes no function how does it function as a calculator? The purpose of the physical analysis is to understand the necessary relation between those physical states we have designated as inputs and those outputs.
DeleteRoss addresses that very point in the article: machines don’t “add” or “fly” — they simulate addition or flight; that is, they act in a way that bears some useful correlation to addition or flying. Just like streaks of ink don’t “mean” anything, human beings mean things, by way of inky streaks; and human beings add, sometimes by means of a machine. Of course, it’s a convenient and obvious shorthand or metaphor or analogy to say, “these squiggles mean C-A-T” or “this machine performs the function of addition”, but the meaning or function requires a human; the physics alone (ink, cogs and levers, etc.) is not enough. As soon as you “designate” something as an “input” or “output”, you are going beyond the physics and adding an interpretation.
DeleteWe have no access to any underlying medium of thought which we might have independent reason for thinking reliable. Such as matter.
I don’t see how any doubt is cast on Ross’s argument unless you assume that the mind is material, and if you go in with the assumption that Ross is wrong, it’s hardly surprising that you conclude his argument doesn’t make sense. In fact, Ross also directly addresses Kripke’s conclusion: he says that, 'Kripke drew a skeptical conclusion from such facts, that it is indeterminate which function the machine satisfies, and thus "there is no fact of the matter" as to whether it adds or not. He ought to conclude, instead, that it is not adding'.
Given Kripke’s demonstration that, as you put it, we have no reliable medium of thought such as matter, then either we don’t really have thoughts, or else we must have a reliable “medium” that is not matter. Ross’s argument is simply that we do have thoughts (of the appropriate determinate kind), therefore the mind is not material.
By the way, I'd like some thoughts on that:
DeleteI am under the impression that, whether Ross's argument about indeterminacy works or not, the intentionality of thoughts is itself something that cannot be characteristic of matter, even if we assume an Aristotelian conception of matter.
This is because the immanent teleology of natural objects seems vastly different from the "aboutness" of a thought. The match is directed towards fire as a final cause. This can be easily understood as the match having properties with causal dispositions to produce fire. My thought about an individual person, by contrast, seems to have nothing to do with causal dispositions - rather it is like I have the person in my thought, but not just as a representation (since my thought is about the person, not any picture of it, say). I can understand how physical properties could lead a match to be "about" fire, that is, disposed to specifically cause fire. I can sorta understand how a dog eagerly anticipates a meal by having physical properties of the meal interacting with its senses. I cannot make the same sense of a thought about a person.
It seems all these other "intentional" states (match being causally disposed towards fire, dog anticipating a meal, plant being directed to gathering nutrients, etc) are not representational. They involve good old causal relations of a familiar kind.
But the intentionality of thoughts is "representational". When I think of Edward Feser, I don't have a causal power to produce him, and my body is not just reacting to him. I just am grasping him somehow, representing him. It's hard to describe well since it seems so basic.
But then the same issue of intentionality remains. Just as words and ink marks on paper do not by themselves mean anything, how can our neurons or physical relations with objects "mean" anything? The familiar problem of cognitive intentionality seems to me more basic than the issue of determinacy versus indeterminacy. And Aristotelianism doesn't change it at all; the cases of immanent teleology and "aboutness" we find in nature are very different in kind from the aboutness of our thoughts. In a sense, our thoughts (or our interpretations of representations) just are what they are about.
We Brits have a hypertrophied sense of irony :-).
DeleteBut I fear nobody is addressing my objection. Ross says,
Whatever the discriminable features of a physical process may be, there will always be a pair of incompatible predicates, each as empirically adequate as the other, to name a function the exhibited data or process "satisfies."
For instance, there are no physical features by which an adding machine, whether it is an old mechanical "gear" machine or a hand calculator or a full computer, can exclude its satisfying a function incompatible with addition, say, quaddition.
The consequence is that a physical process is really indeterminate among incompatible abstract functions.
These claims are all inconsistent with what we might call the Engineer's intuition: On the finite domain of inputs on which the machine produces a result a physical analysis of its operation reveals that the relation between inputs and outputs is uniquely determined, subject to caveats about standard operating conditions, of course.
Is the Engineer wrong then? Is he using words differently from Ross?
I’m not really sure what it is that wasn’t addressed in previous comments or by my quotations from Ross; but you're cheating by switching metaphors mid-sentence: didn’t we agree that a “physical analysis” does not include “inputs and outputs”? The physicist doesn’t see inputs or results or functions; he sees what is happening — particles in motion. The engineer sees what is supposed to be happening — he’s read the specs in the manual and supposes, "puts upon" an interpretation: these particles are considered to make up an input, those particles an output, etc. But the physicist will never turn into an engineer simply by staring at particles.
DeleteAtno: The familiar problem of cognitive intentionality seems to me more basic than the issue of determinacy versus indeterminacy.
DeleteYes, certainly; the determinacy is just a way to prove that there’s no possible objection along the lines of, “Maybe that intentionality is an illusion caused by something that exactly corresponds to the object of thought in a physical way”.
@Mr. Green
DeleteDoes that also apply to Dretske and other naturalists who attempt to explain intentionality in a naturalistic way via natural signs and associations as forming intentionality?
Because if matter is intrinsically meaningless on it's own, it's hard to see how mere natural causal consequences or connected associations can somehow lead to true intentionality. Just because a doorbell is often caused by a living being, doesn't mean that the doorbell somehow points to a human being, anymore than combining the word "orange" with an orange beside it somehow generates intentionality - both are meaningless in themselves, and also when combined together.
No matter how many 0's you add, even if it were infinitely many, you will never get more than 0.
Is that correct?
@JoeD
DeleteI think that is a real problem. To make it even more obvious, just look at how completely different the following cases of "intentionality" are:
A man thinking about you.
The arrow in a compass pointing towards North.
A dog tracking a fox through its smell.
A plant needing nutrients to grow.
A match producing fire.
If you look at this list like I do, you'll notice that the first case (a man thinking about you) is almost NOTHING like the other examples. And mind you, the man isn't even thinking of a universal concept, he's just thinking about you, an individual. The "aboutness" of a thought seems entirely different from the derivative intentionality of compasses and other artifacts, but also entirely different from the intentionality of a match producing fire or a plant needing nutrients or a dog tracking a fox. It's completely different.
The problem is I find it hard to communicate clearly what the issue really is. It's almost as if the intentionality of thought is ineffable; it certainly is very difficult to describe it in a more basic way, we just feel tempted to repeat ourselves: it just is thinking about you; it just is to have you in the mind; etc.
But notice how the other examples of intentionality (including those favored by Aristotelian immanent teleology) can in principle be explained through physical processes. The dog tracking the fox, the plant needing nutrients, the match producing fire, all can be traced to causal dispositions and powers in properties, or physical reactions and relations. The dog is "tracking the fox" by actually smelling whatever vestiges the physical body of the fox left behind; the plant needing nutrients just refers to the plant's inner constitution and physical properties having some specific causal dispositions, including a disposition to react a certain way to nutrients; the match simply has properties with a causal disposition to produce fire, etc.
But how about thinking about you or any other individual? What is that? By virtue of what physical property is a thought "about" anything at all?
Notice how it seems like only thought is representational. All other examples can be given a description or explanation that does not involve representing the intended object.
It's too bad it's so hard to speak clearly about intentionality, given how basic it is.
@Atno,
DeleteI wouldn't be so sure about the dog tracking fox case. Aquinas himself acknowledge animals have intentions (bird searching branches to build a nest), and the other animal examples I cited seem to show some form of non-rational intentionality in animals.
I am completely sure about the dog tracking fox case. It's rather like the case of me tracking the location of a food plate in my house - my sense of smell is much worse than that of a dog, but I can still track some stuff with it. Just like I can still tell whether there is gas in my kitchen just by smelling gas (and I can also kinda tell where the kitchen is, by smelling the gas from afar). And let me tell you, it is nothing like when I am thinking about my particular cat.
DeleteI can make sense of "tracking intentionality" by reducing it to causal relations between the olfactory senses and the properties that the dog is smelling. Aside from the hard problem of consciousness (which I don't think Aristotelianism solves, by the way), there isn't much of a problem there. I can't do the same with my thought about my cat, however. My cat as a whole individual is, somehow, literally in my mind - my cat, really, not properties of my cat or anything like that.
Thinking about a fox seems to me completely, utterly different from tracking a fox by smelling its trail of smelly properties.
I am not convinced animals exhibit this kind of intentionality. I don't think they "think about" particular objects; they seem to me to merely react and interact with qualia, including in very ingenious ways. But they don't think about things.
(If they do, however, then here might lie a division between rational animals: humans are rational animals that can think of universal abstract concepts, such as humanity; dolphins are rational animals that cannot think of universal abstract concepts, but can think of particular objects, such as humans. Some animals can't think about anything all).
@Atno,
DeleteOr maybe rationality isn't simply having intentionality, but knowing in the mode of universals? There may be multiple levels of immateriality, of which only rationality is fully subsistently immaterial.
Mr G, I claim that the physicist does see inputs and outputs. He calls them the initial and the final states of the evolution of a dynamical system. His dynamical analysis then reveals that there is a unique functional relation between inputs and outputs.
DeleteI don't believe I'm using metaphorical language at all---you'll have to show me where I am mixing metaphors.
What we agreed back at the beginning was that we were picking out for special interest only certain physical aspects of the device, that these aspects could be usefully distinguished one from another by associating them with numbers, and that none of this depended on a theory of mind. I'm afraid I don't see where meanings and interpretations come into this at all.
Well, in real life a physicist can always put on his engineer’s cap, of course. But for the purposes of our example, our physicist sees nothing but physics — that is, all he can see are atoms and their interactions. How can he tell which atoms are the alleged “inputs” and which are the so-called “outputs”? (We’re not talking about the inputs of his equations for describing the atomic system, after all.) Do the input-atoms follow different laws of physics from the outputs? Are they labelled somehow? (And what could they possibly be labelled with? More atoms?)
DeleteNow a “function” is not merely anything that a machine happens to do, but some representation of an abstract operation (such “addition” or “quaddition” or “modus ponens”). So I present to a you big pile of atoms stuck together: you may now proceed to examine away, and let me know what the function is, where the inputs and outputs are (if any), and so on. What experiments are you going to perform to answer these questions?
What part or aspect of the system counts as an input or output has to be decided by us in advance. I said this back at the start. Obviously, by choosing different parts we may find a different relation between input and output. So, yes, there is an indeterminacy here. But this applies to Ross too. He is saying, or so it seems to me, that even when what counts as an input or output is decided, and this indeterminacy is resolved, even then the relation between input and output is indeterminate. That, I think, is quite mistaken. So, contrary to what Ed and others aver, Ross's argument adds nothing new to compel us to the conclusion that mind is immaterial.
DeleteWhere is Ross saying that? And even if he did, how does that affect the argument?
DeleteAs quoted above:
DeleteWhatever the discriminable features of a physical process may be, there will always be a pair of incompatible predicates, each as empirically adequate as the other, to name a function the exhibited data or process "satisfies."
My bold. It's clear what counts as data has been decided. It's clear also that Ross is restricting himself to considering finite histories of inputs and outputs, just as Kripke does. But Kripke is talking about what we can know of our own minds. We can't open up our minds to look inside for evidence of their reliability. That's what pushes us towards a sceptical conclusion. But with an inanimate physical object we can do physics on what we find inside.
Thanks, David, I see what you mean now. Forget “infinite” functions like addition and quaddition; consider instead the very “finite” function potrzebie, according to which 0•1 = 1, 1•0 = 1, and that’s it. Given suitable definitions for what counts as an input and an output, and as “0” and as “1”, we can build a machine that performs exactly that function. It also happens to overlap with part of addition (and quaddition, and infinitely many other possible functions); but it only exactly determines potrzebie (anything else exactly equivalent to that would just be potrzebie under a different name). Right?
DeleteI don’t think this means Ross’s argument is wrong, although he could have been more explicit about the details. I think he is talking about inputs, etc. for convenience, as though we could simply help ourselves to these notions, just so he can discuss Kripke’s example on the same terms as Kripke does. The fact that even that much interpretation is beyond what can be physically determined just makes Ross’s argument that much stronger: even with that extra layer of interpretation given for free, there are still functions that cannot be realised determinately; all the moreso is it impossible to get any determinacy without that leg-up.
We can't open up our minds to look inside for evidence of their reliability.
Again, this surely just helps Ross. Even if we could help ourselves to a free interpretation of inputs/outputs/etc., and even if we could take this interpretation and look inside our brains, we still couldn’t get the necessary determinacy. Even if, per impossibile, we could do all that, it still wouldn’t provide a way out because we do not in fact hold determinate thoughts in mind by means of cracking open our brains and checking their wiring first.
Yes. And in the set-theoretical picture of functions the idea of one function being part of another or overlapping another can be made precise.
DeleteThere is a long standing argument that Ed has rehearsed several times to the effect that physical representations are indeterminate in themselves as to what they represent. This pushes us to the conclusion that mind is immaterial. It's a strong argument. There is an analogue to this in the present discussion: the calculating device in itself is indeterminate as to what it calculates. This too is a strong argument. Ross thinks that his Kripke inspired argument reveals a yet deeper level of indeterminacy in the physical. I have been arguing all along that he is mistaken in this. If I'm right then Ross's work doesn't add anything to the long standing indeterminacy of representation argument. He doesn't give us any further reason for thinking that mind must be immaterial. That is all.
Further thoughts?
DeleteDavid, I think you might be right that the argument overall is equivalent to the general representation version. I don’t know whether Ross thought otherwise, or just that this was a particularly precise way to formulate the argument. Having pondered it bit, though, I think Ross’s version may have an advantage against someone who bites the bullet and admits that a brain can’t, say, determinately represent a perfect triangle — but claims that we don’t need our brains to do that, as long as we can get a “close enough” approximation. Ross doesn’t simply say, “Of course you can conceive of perfect triangularity”; he says, “If you can’t determinately perform functions, then you can’t do basic logic like modus ponens, so you don’t have an argument at all.”
DeleteWell, perhaps. But I think it concedes too much to Ross to accept what appears to be his view that a material mind must work through analogues, either exact (Ross) or approximate (his opponent).
DeleteI looked at Ed's ACPQ article. The nearest he comes to answering my objection is on page 21 where he says,
...Dillard says that there is a determinate difference between an and-gate, an or-gate, and other logic gates, which falsifies Ross’s claim that physical phenomena are inherently indeterminate. But this simply ignores Kripke’s point that whether a machine has certain computational properties—in this case, whether a given electrical circuit really instantiates an and-gate or is instead malfunctioning—is not something that can be read off from the physical properties of the circuit itself, but depends on the intentions of the designer.
I'm not a trained philosopher so I don't have to be in awe of Kripke. Does he say this in so many words? My view is that this is diametrically wrong. Whether a circuit is an and-gate depends solely on its behaviour which itself depends on its physical structure. The intentions of the designer are irrelevant. After all, he may have wanted to build an or-gate but got the designs mixed up. But I'll look at the Kripke original.
Kripke discusses machines on pages 32--35 of WRPL. Machines come up because one anti-sceptical proposal for stabilizing an interpretation of 'plus' is to externalise it in the form of an adding machine. Kripke argues that this cannot defeat the sceptic. He says (p34) there are three problems.
Delete1. There will have to be rules to be followed for setting up the inputs and interpreting the outputs and these will be susceptible to the sceptical attack.
2. The machine is finite and cannot perform additions on arbitrarily large numbers.
3. Machines malfunction. The hoped for stabilization is a mirage.
Problems (1) and (2) are not relevant to the question of the determinacy of a physical machine. Strictly speaking (2) is a non-sequitur. We can envisage a serial adding machine that receives the digits of the summands in pairs, least significant pair first, and outputs the corresponding digit of the sum, just as we were taught to add as children. Too much is made of problem (3). We are rather good at engineering machines that perform stably across the range of conditions found near the surface of the Earth, as is nature. And as far as we can tell our minds are far from perfect too. Most of them work for just two thirds of the time, some never get to work properly, and all seem subject to distraction and caprice and error, and all ultimately wear out.
Regarding Ed's summary of 'Kripke's point' in the above quoted passage, what Kripke appears to say is (p34),
The machine as physical object is of value only if the intended function can somehow be read off from the physical object alone.
This is in the context of problem (2) where
Indefinitely many programs extend the actual finite behaviour of the machine.
But what has this to do with the indeterminacy of the physical machine? The machine is quite determinate. What appears to be indeterminate from the machine alone is the extended function the designer may have had in mind as he worked. Sure, the intentions of the designer can only be guessed at, but what the machine actually does is a question of physics. Ed seems to think he can dismiss Dillard with a Kripkean thought taken out of context.
But I think it concedes too much to Ross to accept what appears to be his view that a material mind must work through analogues, either exact (Ross) or approximate (his opponent).
DeleteI’m not sure what you mean by “analogues”. If you mean it in the sense that what is in the brain is analogous to the actual concept [the form, in the Aristotelian sense], then what alternative is there? It can’t be the concept [in the sense of the form’s being instantiated in matter], because concepts clearly are not physically instantiated inside anyone’s brain.
Whether a circuit is an and-gate depends solely on its behaviour which itself depends on its physical structure.
[…] 1. There will have to be rules to be followed for setting up the inputs and interpreting the outputs and these will be susceptible to the sceptical attack.
Right, this just brings us back to having to define what counts as an input, etc. The physics is determinate, but we can’t get from there to an equally determinate meaning.
2. The machine is finite and cannot perform additions on arbitrarily large numbers.
I’m not sure how much this matters if the point is supposed to be a (somewhat functional/behavioural) approach where we are allowed only to use the machine from the outside (insofar as the machines are stand-ins for our minds, and we cannot crack open our minds to see how they work). But Kripke points out that a finite machine can, in fact, add arbitrarily large numbers; he also points out that this doesn’t work for multiplying. (His original example probably should have used schmultiplcation… oh, well.)
3. Machines malfunction. The hoped for stabilization is a mirage.
And as far as we can tell our minds are far from perfect too.
Sure, but the question isn’t whether we can ever get things wrong, but whether we can ever get things right. The argument is that matter-only could never do better than the accidental correctness of a stopped clock. (Indeed, on an Aristotelian view, our thinking can go wrong only because it’s tied to a physical brain. Angels cannot make mistakes.)
What appears to be indeterminate from the machine alone is the extended function the designer may have had in mind as he worked.
Which is the only thing we care about (in this context). If the designer can’t get the extended function from his mind into the machine, then he can’t get it into his brain, either.
Mr G, I think you misunderstand what I'm trying to do here. I'm not attacking Ross's conclusion that mind is immaterial or whatever. For all I know this is right. Even less am I attacking Kripke in WRPL. What I am trying to show is that the Ross/Feser argument for this conclusion doesn't hold water. It doesn't give us any further justification for the conclusion than its predecessor arguments.
DeleteIn particular, Ed hasn't at all refuted Dillard with his Kripkean hammer. Kripke says that from the machine alone we can't infer its designer's intentions. Fine. How is this supposed to rebut Dillard's point (and my point) that machines can behave determinately?
>What I am trying to show is that the Ross/Feser argument for this conclusion doesn't hold water.
DeleteYou might be right that it doesn’t add anything — but that doesn’t mean the argument doesn’t hold water, merely that it doesn’t hold any additional water. The Ross/Feser version is a bit more complicated than it needs to be, that’s all. And if that’s true, Dillard’s objection is already refuted by the simpler version of the argument. You and Dillard are quite right that the machine is determinate in what it does physically; but nobody cares about that, because atoms can’t add (or schmultiply).
We can get even simpler: forget the machines and look at some words printed on a page. What do the words mean? You cannot tell me. It’s indeterminate. Dillard can insist that he’s studied the page under a microscope and has a completely determinate, exhaustive description of the ink and wood-pulp. But he still can’t determine what the words mean. There are different things involved, at different levels, and it doesn’t matter how determinate things are a one level if they are still indeterminate at another.
Doesn't hold water was intended as a polite way of saying is invalid. I can make no sense of additional validity. You and I agree with Dillard that the machine is physically determinate. Ross appears to argue otherwise. His error, in my view, is to consider only histories of input output pairings and to ignore the possibility of investigating and understanding its physical structure. I don't think Ed has addressed this at all.
DeleteI don't think Ross is about meanings. The word doesn't appear at all in Immaterial Aspects of Thought.
If your point is that Ross was misunderstanding something, well, I dunno what was going on inside his head, and I have to admit that I don’t particularly care. I’m interested only in what the best version of the argument can show; and if Ross is unclear, that’s why Feser wrote his paper. And Feser does quite explicitly refer to meaning.
DeleteWell, let's take a closer look at what Ed says. On page 16 in his account of Ross's argument we find,
DeleteRoss, again following Kripke, notes that there are no physical features of an adding machine, calculator, or computer that can determine whether it is carrying out addition or quaddition, no matter how far we extend its outputs. As Kripke emphasized, appealing to the intentions of the programmer will not solve the problem, because that just raises the question of whether the programmer really had addition or quaddition in mind, as in the original paradox. But Kripke makes a deeper point. 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. Now 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 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. We might find out by asking the programmer, but there is nothing in the physical properties of the machine itself that can tell us.
The emboldened claim is mistaken. Ed follows Kripke in drawing a distinction between the physical machine---presumably some kind of stored program computer---and its program. Fair enough. Program sounds abstract and conceptual, machine sounds concrete and physical. But to turn the bare machine into an pluser or quuser the program has to be made concrete and physical as a pattern of charges loaded into the machine's memory cells. This distribution of charge forms part of the initial state of a prepared dynamical system governed by the laws of physics. Even if we are ignorant of the process of writing, compiling, and loading the program, this initial state can be recovered, and in combination with knowledge of the structure of the bare machine, we can tell whether the prepared machine will plus or quus. So there is something in the physical properties of the prepared machine that will tell us what it will do. What it doesn't give us is a determinate understanding of the programmer's intentions. Maybe he intended to make a quuser but forgot to include the critical test on the magnitude of the inputs and ended up with a pluser.
The emboldened claim is not mistaken. (Nor is Ed drawing any distinction between hardware and software.) I think you are getting too hung up on the particular example of quadding. Sure, given machine X with some particular Kripke-indeterminacy, it may be possible for you to build a different machine, Y, which does not suffer from that particular problem. That does not defeat Ross/Kripke/Feser, though — machine Y will have some other problem.
DeleteMaybe he intended to make a quuser but forgot to include the critical test on the magnitude of the inputs and ended up with a pluser.
Yes, the “broken quus” function can implemented in exactly the same way as the “working plus” function. And the same way as infinitely many other possible functions. You can tell exactly what the machine is doing physically, but you cannot tell which function it is performing. (Or, if you insist, you can say it is performing infinitely many functions all at once.) That is the point of the argument: the mechanism is determinate, but the functionality is not. There is no one-to-one mapping between them. Any mechanism implements infinitely many functions, and any function can be implemented by infinitely many different machines. Therefore “functionality” and “mechanism” are two different things.
Not sure what you mean by a particular Kripke indeterminacy. As far as I can see from WRPL, Kripke's argument is that if I make a machine to 'capture' what I think 'plus' means then my intention is undetermined by said machine. This is very general and applies to all machines. I accept this. I agree with Kripke that this gambit doesn't defeat the sceptic. Ross wants to go further and show that there is something indeterminate about the machine itself, and here I disagree. I think that the argument I gave in my last comment defeats Ross's claim. I have said that Ross's mistake is to ignore the possibility of 'taking the machine apart' to see how it works. If you think my argument is faulty please tell me where I go wrong.
DeleteYou say that we can tell what the machine is doing physically but we can't tell what function it's performing. Again, I'm not sure in what sense you are using the word 'function'. I confess that Ross also sometimes uses 'function' somewhat mysteriously for me. Since we are talking about mathematical functions I take it to have its usual mathematical meaning, namely, the relation between the input pairs and the outputs. And this I say is fully determinate, once we have decided what counts as input and output. If you then say that we can regard almost any aspect of the physical machine as an input or output then I say that Ross's argument also depends on deciding in advance what counts as input and output. If you then rest your case on this freedom to choose then I reply that Ross has added nothing with his use of Kripke. We have been here before.
then I reply that Ross has added nothing with his use of Kripke. We have been here before.
DeleteIf Ross has added nothing, then he is not wrong, just extravagant. (A steamroller is not necessary to crack a walnut, but it will indeed split the shell.) If Ross is right — even if he’s driving a steamroller — the the aforementioned emboldened claim is true. So you seem to be making two claims: the argument in Ross’s form goes overboard (which may be the case), and that the argument in Feser’s form is wrong. But if Feser is wrong then Ross is not “extravagant”, he too is just plain wrong. We have been here before, but I get the feeling that each time we get back here you go off in the other direction.
Now, I also wish Ross had defined the term “function”. In context, he certainly means more than just a mapping; I’d say his definition has to incorporate meaning or intention. (For instance, if f(x) = x+0 and g(x) = x×1, then the mapping of ℝ→ℝ is the same for both functions, but clearly adding and multiplying are different operations.) Perhaps this is leading to confusion over the word “machine”. Really, a machine is a device for performing some task — some Rossian “function”. It’s not merely a pile of parts (mere physics); the term implies “parts organised to perform some specific actions”. So it’s true in some sense that we can read the function off from the machine, because it’s only a “machine” in virtue of already knowing something about the operation(s) it is supposed to perform. So we can read (perhaps, some) determinacy off a machine but not off the mere clump of atoms that it constitutes on a purely physical level.
And this I say is fully determinate, once we have decided what counts as input and output.
OK, here’s my machine: it has two inputs, buttons that you can push in or out, and an output, a light that turns on or off. When either one button or the other is depressed, the light goes on. When both buttons are depressed, or neither, the light goes out. You can ask any question about the physics (but there’s nothing else to discover other than that the machine behaves exactly as I’ve described.) So you can tell me what function the machine is performing, right?
Ross is wrong in so far as his use of Kripke to demonstrate the indeterminacy of the physical falls short of its goal. The jewel, or rather Ross's use of it, is flawed.
DeleteI fail to see a distinction between the machine and the formed matter that constitutes it.
The relation between inputs and outputs is what computer people call 'exclusive OR' : (0,0)-->0, (0,1)-->1, (1,0)-->1, (1,1)-->0. The exact way of achieving this doesn't matter as long as it's reliable. The circuit an electrician would install for controlling a hall light from two switches would do.
And that failure means you failed to tell me what function the machine serves. You see, it is actually my “The British are Coming” machine: one button if by land, two if by sea, and a light for signalling Paul Revere. Of course, now that you mention it, it occurs to me that the same machine could by a strange coincidence ALSO serve perfectly well to perform the XOR function. And now that I’m mentioning it, it will occur to you that it could also equally well serve to perform the XNOR function (which are of course two different functions — in fact they are opposites).
DeleteYet you clearly understood exactly what the machine was doing physically. You have yourself demonstrated Ross’s conclusion: knowing the physics, or even that it was a machine, or even what the inputs and outputs were, you were unable to determine “the” function. Physics simply cannot determine whether it is “really” performing the British-are-coming function or the XOR function or the XNOR function or the paperweight function or any other that might apply.
In the case of the PR machine I think you are equivocating on 'function'. It's clear that Ross is talking about mathematical functions, not purposes.
DeleteIf we take 0 to represent light off and 1 to represent light on we get XOR. If we take 0 to represent light on and 1 to represent light off we get XNOR. But we agreed earlier that what counted as input and output and how these physical states were to be represented by numbers had to be decided in advance, else Ross's use of Kripke's quus can't get started. His claim is that even after all this is decided the function is indeterminate. Not so.
A few comments ago, what Ross meant by “function” was “mysterious”. Now that you want to attack him from a different angle, it is suddenly “clear” to you that he means exactly what you mean by “function”.
DeleteWe agreed — or so I thought you confirmed — that meanings had to be decided in order to get a particular degree of determinacy. A physical pile of parts does not tell you there is a machine. Agreeing that there is a machine does not tell you the inputs and outputs. Agreeing on the inputs does not tell you the meaning. Agreeing on the meaning does not tell you whether the machine is malfunctioning or was assembled incorrectly, etc. The are multiple levels of indeterminacy, and that we can’t even get partial determinacy without making these agreements just proves Ross’s point. Of course we can get determinacy after we agree on such issues; you couldn’t determine XOR vs XNOR because we hadn’t agreed on enough. But we agreed completely on the physics; therefore the physics is not sufficiently determinate.
Well, in several places Ross talks about 'pure functions', and I'm not sure what he means by this. But the main plank of his argument is his use of Kripke's quus and Kripke is definitely talking about functions in the mathematical sense.
DeleteForget about meanings! Ross doesn't mention them. We are using numbers to label or name certain physical states. We get the first input into the state labelled '2' and the second in the state '3', press the start button, the machine whirs for a bit and stops with the output in the state labelled '5', and so on. Given these assignments of numbers to states, saying that the machine adds is a very compact way of describing what the machine does. The physics, in other words. And we are agreed on the physics, it seems. No mention of meanings.
For someone who is talking about functions in the “mathematical sense”, Kripke sure uses the word “meant” a lot. Ross uses it too, in reference to Kripke’s argument; although he uses the word “understanding” a lot more, as well as a mention of synonyms (“Equivalent but nonsynonymous functions would give the same arrays from inputs to outputs”). Sure sounds to me as though he means something more than just mathematical mappings.
DeleteBut you insist that he is talking about only physics — if that were true, then yes, his argument certainly fails. Saying that physics can’t determine the physics is not just a failed argument, but a stupid one; and since I don’t think Ross was a moron, I believe your interpretation is a pretty uncharitable reading. I think it’s much more reasonable to see him as saying something about semantics or interpretation or intentionality. However I don’t have anything more to say about that way of reading him that I haven’t already said.
OK, let's leave it there.
DeleteThere is no crisis in Physics. Physics is doing well thank you. The crisis is in the minds some people who do not get the idea. The reason for this is that it has gotten hard and takes longer to get the idea.
ReplyDeleteQM works as well as when it was started. So does Relativity. There were problems with QM in terms of interactions so Q field theory was worked out by Feynman and Schwinger. There were strange kinds of relations between mass and spin so strings were suggested. [spin seems to increase with mass squared. Regge trajectories.] Gravity was a problem and it turned out that strings have answer to the problem of gravity.
"There is no crisis in Physics"
DeleteWell obviously there is at least one physicist who disagrees. And it shouldn´t be hard to find more.
What exactly did you disagree with Hossenfelder about? She demands change in the method of pursuing physics due to its stagnation for 80 years now.
As a physicist, I affirm: there is a VERY MAJOR crisis in theoretical physics; we really haven't advanced a single inch in these past couple generations. And indeed, it speaks volumes that today we've only got two types of physicists: string theorists, for whom string theory is the be-all end-all in the quest for the so-called theory-of-everything, and the ones who laugh at them (I'll let you guess which ones comprise the vast majority of physicists, even though pop science gives the opposite impression).
DeleteSee, we already have two- Do we get three?
Delete(I'm the previous anonymous)
DeleteBTW, I want to clarify that when I say we haven't progressed one inch I don't mean that most physicists' work is completely worthless (I'm one of them after all). Admittedly, substantive work has been achieved on the more mathematical side of the picture, and certainly on the experimental and engineering fronts as well.
What I mean is that we are nearly as clueless today as we were in the 1960s regarding the direction we should follow in order to make quantum mechanics and general relativity fully compatible and to unify all four fundamental interactions. But the only reason I say 'nearly' is because at least some attempts have been made... even though nobody has any idea of how they might even be tested, let alone whether they are anywhere close to being correct. String theory looked quite promising in the 1980s, but today only those who have dedicated their entire careers to it (and thus have everything to lose by giving up on it) seem to still place in it any sort of hope.
All in all, I agree with Smolin and Hossenfelder's overall assessments. And I think a return to the basics is in order, because: 1) 99.9% of physicists (or scientists in general) know absolutely squat about the history and philosophy of their own field (that is, they don't even know that many of the ideas they take for granted directly rest on philosophical precepts they don't even realise they have and which, rather contingently, were introduced in some specific historical context to tackle some particular concern of some random individual); 2) most graduate textbooks of every subfield are a complete mess (and their lack of mathematical rigour makes mathematicians puke), which is a problem when the subject at hand is already pretty damn hard; 3) the pressure of the 'publish or perish' mentality stops people from thinking deeply on foundational issues and makes them instead get on with business as usual, i.e. publish loads of useless papers that no one is ever going to read (even if they cite them).
In the end, A-T will save the day once again.
Anon,
DeleteThe rising interest in A-T philosophy among some scientists and philosophers of science is a ray of hope. Have you personally seen any signs of it?
Certainly when it comes to Aristotelianism, not so much of Thomism as of now.
DeleteDavid,
DeleteFrom my personal experience, the vast, vast majority of physicists are complete philistines: absolutely clueless about the foundational philosophical aspects of their discipline, and proud of it. In this respect at least, Hawking, deGrasse Tyson, etc. indeed appear to be representative of the whole bunch, unfortunately.
Now, amongst committed Catholics (which are, as you may suppose, a very small minority, especially within academia), I do see some people (though by no means all) starting to read and discuss Dr. Feser's work.
Considering linguistics is broadly mentioned here, I have a question regarding so-called animal language. Namely, in the 80's and 90's a marine biologist Louis Herman attempted to test whether or not dolphins have language capabilities.
ReplyDeleteFor example, he taught them gestures for things, such as Pipe, and gestures for actions such as Touch and Over (swim over). He then tested them to see if they would understand a combination of those, such as Pipe Touch, and the dolphins instantly knew to touch a pipe.
He then tried to see if they could also comprehend conjuctive sentences such as "Pipe Touch Pipe Over", and the dolphins instantly knew to first touch and then swim over a pipe. They then claimed that, because conjunctions are necessarily formed via recursion, and because the dolphins can follow "conjunctive sentences" such as Pipe Touch Pipe Over, that they therefore are capable of and understand recursion!
They are also impressed by the dolphins ability to generalize objects - i.e. they were taught the sign for Hoop which was first an octagonally shaped hoop, but they also applied this to square and circular hoops also, and to hoops of any color, as well as hoops that didn't float but sank to the bottom. Drawing from this the conclusion that dolphins understand the gestures to refer not to particular objects but to whole classes of objects, thereby showing they have "abstract concepts"!
And they make a big deal of the fact the dolphins can perform a lot of tasks even after a delay. For example, the dolphins were given a command requiring a certain object, but would delay introducting a whole bunch of objects (including the commanded one) into the tank for a certain time period. There were delays of 7, 15, and 30 secs, and 15 tasks were tested for each delay. Resulting in 1 mistake in 15 for 7 and 15 sec delay, and 5 mistakes in 15 for the 30 sec delay. They conclude from this that dolphins understand the commands to refer to things across space and time.
What do you think?
Joe, I would think the same thing I always think — if a machine can do it, it can hardly prove intelligence. And my computer is more “linguistically” advanced than a dolphin could ever dream of. Including being able to do actual recursion, not mere conjunction (which is formed by conjoining, not by recursion — at least, the example you gave doesn’t have any recursion in the sense of nested structures; it’s just two commands, one after the other).
DeleteThe other thing I always think is that an example is not very impressive if all it shows is something known to Aquinas a millennium ago, or to Aristotle a millennium before that. Nobody is surprised that animals can be trained to perform tricks, or that they can remember things, or that they can recognise generalities.
@Mr. Green,
DeleteA research paper I read literally stated that conjunctive sentences are important because they are related to recursion / come about by recursion, and go on to say this is proof dolphins cano do / understand recursion, so you shouldn't underestimate their linguistic capabilities.
Believe it or not.
@Mr. Green,
DeleteHere is a book on google books that talks about this example:
https://books.google.hr/books?id=bJ1oAgAAQBAJ&pg=PA147&lpg=PA147&dq=PIPE+TOUCH+PIPE+OVER,+dolphins&source=bl&ots=0nYm2rYzKA&sig=ACfU3U20b5EYJAjLJ0-rMoX5z97GB_Q2eA&hl=hr&sa=X&ved=2ahUKEwi19am9s-rmAhWYw8QBHS-HBFoQ6AEwC3oECAYQAQ#v=onepage&q=PIPE%20TOUCH%20PIPE%20OVER%2C%20dolphins&f=false
And even that book points out that this isn't really true recursion, but rather an iterative tail recursion.
The book also mentions an example of a dolphin combining the whistle sounds for "ball" and "ring" into a single unique sound when playing with both of these objects, thereby seeming to show novelty generation or making new words for new things.
Joe,
DeleteI think you missed the general thrust of Mr. Green's reply. His point is that nothing in the study significantly alters many previous thinkers', especially prominent philosophers', views on animals with higher cognition or mental capacity. So I'm a little confused by your responses.
Moreover, drawing the type of conclusions you seem to from one series of (potentially?) un-replicated study is dubious. While the studies may show what they purport, there is also a chance their results are meretricious. I'm always cautious when it comes to research in which the scientists involved appear to approach the subject with preconceived notions that they then proceed to read into their results. Failing to completely eliminate or limit some elements of bias from an experiment (especially in biology) is one thing, but fitting results into a specific interpretive framework is another. I cannot say with certainty that is what happened with this study or series of studies, but one should always withhold judgement until further research corroborates these findings.
Similarly, one should bear in mind the artificial nature of such experiments. I don't know if there are studies suggesting dolphins express/manifest these linguistic abilities "in the wild," so to speak, but the imposition of particular circumstances by human minds upon research is often read out of the results in these types of experiments as if it is non-existent. It's good to be wary of biological studies implying their findings "prove" something, even though it required a synthetic environment or unnatural mechanism to do so.
Finally, when it comes to "non-rational" animals, as A-T philosophers might refer to them, the human tendency to anthropomorphize them shouldn't readily be dismissed, particularly in our era. I really think many people, including scientists, unconsciously ascribe personalism (in the Wojtylan sense of the term), intellectual (in the Thomistic sense of the term) capacities, and mental lives similar in richness and complexity to those of humans to various other kinds of animals. This frequently colors the thinking of those conducting and translating such experiments, more so if they possess a specific fondness or affection for animals or the type of animals involved.
None of this directly contradicts, better explains, or refutes the research, nor is it meant to accomplish this task. However, it does present some preliminary questions that ought to be addressed before taking the reported results too seriously or basing further assumptions off of them.
@Anonymous,
DeleteYeah, my last response was a bit rushed. Sorry about that.
As for what you're saying:
"Moreover, drawing the type of conclusions you seem to from one series of (potentially?) un-replicated study is dubious. While the studies may show what they purport, there is also a chance their results are meretricious."
Well, the biologists who did these studies clearly show signs of anthropomorphising the dolphins, at least insofar as they confuse imagination with intellect.
The studies are generally accepted in the field of marine biology, and the biologist in question, Louis Herman, didn't just do one study, but several, in the 1980's and 1990's, on the cognitive and "linguistic" competency of dolphins.
So I would say the results are reliable - though the interpretation may be dubious.
"Similarly, one should bear in mind the artificial nature of such experiments. I don't know if there are studies suggesting dolphins express/manifest these linguistic abilities "in the wild," so to speak, but the imposition of particular circumstances by human minds upon research is often read out of the results in these types of experiments as if it is non-existent."
The general consensus is that dolphins do not use anything like the artificial languages in the wild - the artificial systems were taught them in captivity. Though some researches such as V. M. Janik believe there is "a large potential" for syntactic structure to actually be found in dolphin clicks and pulsed calls, although that structure doesn't need to be mapped onto a semantically / referentially rich communication system that leads to a discrete combinatorial system just like human language is.
There is some evidence that the reportoires of the whistles that the dolphins use might be diverse enough and non-random enough that it does indeed transmit a large amount of information - though that also doesn't entail a discrete combinatorial system.
"I really think many people, including scientists, unconsciously ascribe personalism (in the Wojtylan sense of the term), intellectual (in the Thomistic sense of the term) capacities, and mental lives similar in richness and complexity to those of humans to various other kinds of animals."
That would be correct. The biologists in question literally said that the dolphins ability to learn and combine partuclar sign commands such as "ball" and "touch" together is "concept learning". Others also speak of the dolphins ability to follow the artificial language as if it were awareness of complex abstract concepts. There's clearly a categorical confusion between percepts and concepts, and is a predictable result of the modern collapse of imagination and intellect.
"None of this directly contradicts, better explains, or refutes the research, nor is it meant to accomplish this task. However, it does present some preliminary questions that ought to be addressed before taking the reported results too seriously or basing further assumptions off of them."
Yeah, I'm aware. I just wanted to know / confirm that there is a different way to interpret the feats that various animals such as dolphins can do in a non-intellectualist way.
That is why I also mentioned the dolphins ability to combine two known signs into a new one when playing with the two known things the signs refer to.
I feel like most of these studies tend to fall prey to common misunderstandings of what language and rationality consist in. Some animals have very advanced neural systems and are capable of keeping many different associations between images, events and behaviors. That doesn't really tell us that they are rational or are truly thinking *about* things, let alone universal and determinate concepts. On the contrary, it seems from their normal behavior that they don't and it is only in very specific study contexts that scientists seem to claim they do this and this and that.
DeleteAnd think about Searle's Chinese Room for a second. You can't get any more perfect at language than that. The person in the Chinese Room really does seem to be speaking Chinese, answering all your questions in meaningful ways. But she doesn't speak a word of Chinese. If we can get fooled by the Chinese Room, why can't we get fooled by dolphins or chimps with very good rote and association mechanisms in these experiments?
In any case, if they do somehow manage to have some rational thought, that doesn't change the arguments for the immateriality of the intellect. We'd only have to then conclude that some higher animals also have some immaterial functions and naturally immortal souls. This should make us treat them with more dignity, but still (I suppose) less dignity than humans since they do not have the full range of rationality and potential for rational life that humans do.
Though I don't think animals really have rationality or anything like it. I'm much more convinced that we are rather being fooled by their clever non-rational behavior, much like in the Chinese Room example, or when a computer plays chess against us.
@Atno,
DeletePart 1 out of 2
"I feel like most of these studies tend to fall prey to common misunderstandings of what language and rationality consist in. Some animals have very advanced neural systems and are capable of keeping many different associations between images, events and behaviors."
I agree, though the main discovery in the dolphin experiments that is often lauded is that they seem to recognise word order. To fully clarify what this is, the dolphins were taught various signs to associate objects that can be taken (surboards, persons, balls etc.) actions to be done (touch, fetch) and modifiers such as left and right.
They were taught commands such as "Person Surfboard Fetch", which means to bring the Surfboard to a Person in the tank. But "Surfboard Person Fetch" means to take the person to the surfboard. The first word functions as the destination object or direct object, whilst the second is the indirect object or instrument to be brought to the first thing. The dolphins recognised the difference and acted correctly on the first try.
But the most interesting part is when Left and Right modifiers are used. The dolphin would have two of the same object on it's left and right side, and when given the commands "Left Pipe Hoop Fetch" and "Pipe Left Hoop Fetch" recognised that the sign for Left modified different objects in the two commands, and acted accordingly.
This was taken to show the dolphins knew what grammar to follow, and thus had knowledge of proper word order.
"On the contrary, it seems from their normal behavior that they don't and it is only in very specific study contexts that scientists seem to claim they do this and this and that."
Yes, that is the general consensus. Dolphins don't have language in the wild - though the clicks they use are of sufficient richness that they could actually be using it to communicate information, though that doesn't and likely isn't a discrete combinatorial system, or language.
"And think about Searle's Chinese Room for a second."
Well, one problem with that is that the person IN the Chinese Room already knows a language, and is using a Chinese dictionary to link it to the language he already, so the result is still accomplished partially and necessarily with true knowledge of a language.
@Atno,
DeletePart 2 out of 2.
"We'd only have to then conclude that some higher animals also have some immaterial functions and naturally immortal souls. This should make us treat them with more dignity, but still (I suppose) less dignity than humans since they do not have the full range of rationality and potential for rational life that humans do."
Well, it's important to know that one of the properties that flow from having rationality is being truly culpable of one's actions with regards to morality.
But not even the staunchest animal-rights advocate who believes some animals have language would ever say we should start treating animals as if they were morally culpable in the way we are - that we would have to punish dolphins who attack and thus assault any human being or to punish dolphins who kill the young of other dolphins and rape female dolphins in gangs violently (both of which actually often happen in the wild) as being guilty of rape and infanticide.
So with regards to morality and natural law, the difference between humans and other animals is intuitively known to be so vast that they aren't in the same ethical category as we are.
Now you could say that rationality may come in degrees and kinds, and that even though rationality is required for being in the ethical category of culpability, dolphins and other animals who have a lower degree of rationality simply don't fit that, so there is an important difference in degree there.
Or maybe kind too. Your theory of lower rational souls is very interesting, and there could in that case be differences in kind between lower and higher rational souls in that way, not just one of degree of rationality. I would be very interested to see a Thomist develop this possibility further, at least in theory!
And there is something that may hint at that - even if some animals have language, it would be very simple and concrete and thus not suited for the immaterial and trasncedental. In fact, the more immaterial and transcendental you go (such as ethical culpability, awareness that concepts are themselves immaterial, awareness that eternal truths are themselves eternal, religious belief, story-telling etc) the more obvious it becomes that animals simply don't have that.
Even in cases such as the one I cited of a couple of dolphins literally combining two whistle signs into one when playing with a ball and hoop, thus creating a new sign that is a combination of the previous two, it's all non-natural and rare when it happens (because it's not like dolphins always combine two signs into a new third one when being taught artificial sign or acoustic language), and it's still about something concrete that was prompted about by something concrete.
Plus, the dolphins could just as well have been playing around with signals and syntax, without needing to interpret it as them inventing a new word - even birds can do that, Alex the Parrot once referring to an apple as a banerry, due to it's similarity to a banana and cherry.
@Atno,
DeleteAddendum about transcedental and rationality:
I suppose the way I framed "thus not suited for the immaterial and trasncedental." was a bit off.
You did say that lower forms of rationality could still have some immaterial functions rather than just being still stuck in the concrete, but not the fullness of rationality or much of the higher functions we have. This would mean the rationality of some higher non-human animals is immaterial to some degree, but the immateriality our intellect has is way higher than theirs and distinct in kind due to that.
That's what I should have wrote.
But the point of the Chinese Room is to serve as an illustration for how a machine could in principle display (what seems like) syntax without really possessing any semantic content. Sure, we're the ones who programmed the machine or wrote the book that associates such and such Chinese word with another, but the idea is that the whole process of following these rules and associating such and such symbols with another can be a material or computational process, yet that is (of course) not sufficient for understanding, semantic concepts, and so on.
DeleteThe dolphins could be engaging in something similar. If a Chinese Room could fool us, perhaps so could a dolphin. The "language" which governs the dolphins's reactions and responses, which they might not at all understand as a semantic language, could just be a series of behaviors for associating such and such action with such and such "word" and "order". Who wrote this "language"/series of associations? Might have been a result of natural selection or just the dolphins's brains and their high capacity for rote and correlations between sensible data. Who knows. That possibility should be sufficient to give us pause here.
In any case, I am inclined to think that consciousness is itself an immaterial power. I am not convinced that Aristotelianism successfully "naturalizes" consciousness; it helps to ground qualitative features in formal causes of real substances, but that in itself does not account for how a material substance could have a power to actually see and receive these qualitative forms as a first person experience, as qualia. There seems to be an explanatory gap between physical facts and consciousness, so for all I know consciousness might indeed be immaterial. If that is the case then every animal that exhibits conscious experience might have some kind of immaterial soul.
@Atno,
Delete"The dolphins could be engaging in something similar. If a Chinese Room could fool us, perhaps so could a dolphin."
The interesting thing about this is that dolphins have consciousness / a sensitive soul, whereas machines that could do the Chinese Room don't even have that. So we could in principle have an animal that didn't have intellect, but had the processing power and rule-following of a machine, and that could doubly fool us into thinking it had intellect.
"Who wrote this "language"/series of associations? Might have been a result of natural selection or just the dolphins's brains and their high capacity for rote and correlations between sensible data."
Keep in mind that the sign language and acoustic language the dolphins were taught was artificial and made by us - they don't have it in the wild. At least if that's what you're asking.
But yeah, other considerations (especially those of ethical culpability or a lack thereof) are also important.
And I would be very interested to see whether or not Dr. Feser or someone else finds the idea of different types of immaterial intellect, some of which have some immaterial powers while others have a much greater richness, theoretically possible or not.
As for consciousness, one possible answer could be that first-person experience itself doesn't come from the matter, but rather the form and principle, which is immaterial, while still depending intrinsically on matter completely, even though it's distinct from it.
Yeah it was artificial, but based on physical signs and events. The dolphins have a very good capacity of keeping associations between different things, and this might be sufficient for them to "learn" this "language" without intellectual power, the way the Chinese Room is able to fool us with very good abilities of associating symbols and responses.
DeleteAs far as consciousness goes, I don't see how that works. There is no form apart from the substance, the only being that exists there is the informed substance. If the form has powers that transcend the bodily functions, then we're dealing with something immaterial much like the human soul. And if there is an explanatory gap between bodily functions, physical facts, etc. and consciousness, the power to actually experience qualitative forms as a conscious subject, then the bodily functions do not cause this power. The consciousness might not really be intrinsically dependent on the body.
Couldn't there be a zombie that has all the same physical functions as we do - takes in all the physical and chemical information from the world in a series of reactions, etc. - but is not conscious of anything? Nothing in its physical functions would tell us for sure that it is conscious; even if the quale of redness really exists in an apple as some kind of accidental form, how and why would the body's functions be able to receive and experience this qualitative form as a first-person subjective experience? So, couldn't there be a zombie version of humans? Its form would be different from us, but only in that it has no consciousness. It would still lead to all the physical acts and functions we observe in any normal human being. The difference is the form of humans includes a power for conscious experience, the form of the zombie does not. But then the power of consciousness is entirely additional to physical functions and is properly immaterial.
The mere idea that we have to coax the expressions of intelligence out of dolphins seems to count against it. A dolphin developing as it naturally does should display proper intelligence more often than not. Yet it takes the human involvement to bring it out.
DeleteTo say otherwise would entail concluding that pretty much every dolphin suffers some kind of cognitive disorder or defect.
@Billy,
DeleteWell, a possible counter-argument could be in the form of "You can't coax something out of someone unless it is naturally there in the first place!"
@Billy,
DeleteThat is, naturally there as a developable potency, if not a proper accident.
Tim Hsiao from that Capturing Christianity interview has two interesting papers on the perverted faculty argument. Available on academia.edu here and here.
ReplyDeleteWould really like to read few words by mr Feser on Rosenberg' book.
ReplyDeleteIn terms of the first link I wanted to add that Steven Dutch has a an idea about the criminal justice system in the USA. https://stevedutch.net/Pseudosc/MissBoatLib.htm
ReplyDeleteBut no contents anywhere at all? Rosenberg bites the bullet, committing to the idea that meanings and reference are useful fictions. But this claim makes his own position self-undermining, for how does Rosenberg's commitment to the fictional character of contents square with (2) above? Wouldn't any discourse, including his own, lack truth conditions?
ReplyDeleteRighto: Rosenberg's notion undermines science, including his own science, just as much as it undermines history. If the mind doesn't have content, then it doesn't have content when it's doing science. So throw out science too, buddy.
Yes; Rosenberg has graduated from thinking that we don’t think to providing a narrative that explains why narrative explanations can’t work.
DeleteIsn't this just what physicalism and scientism "is" when played out?
DeleteIn order for the mind to be a physical thing it must not exist at all. At least Rosenburg bites the bullet and affirms this, as stupid as it may be.
You mean, instead of physicalists teaching (without saying directly) ""A implies B implies X implies... blah blah blah... I am stupid." Whereas Rosenburg bites the bullet and says outright "I am stupid".
DeleteAfter reading Kirk Durston, I think Sabine Hossenfelder needs to expand her search for where the problems with contemporary physics lie.
ReplyDelete