This whole fight between realist and “anti-realists” has been going on for too long without any result. I am officially challenging Levi to a philosophical face-off: if you claim to know primary qualities of things, i.e. to know characteristics of things themselves, then declare once and for all exactly how you are able to do so or forever hold your peace!
Hi Mikhail,
I’m afraid that I just don’t have an answer to this question at this point. That is, this is a philosophical question and problem to be worked out and solved, not an answer that I have in hand. My thesis is that we require primary qualities to belong to things-in-themselves for certain knowledge claims to be intelligible. That is, knowledge of primary qualities is a quid juris requirement for certain knowledge claims. Now before you jump all over me, I outlined just why I think this is so in my three posts on Meillassoux, as well as in my more recent posts “Correlationism and Creationism” and “The Basic Paradox– A Brief Note”. In my view, correlationist approaches to certain knowledge claims end up falling into absurdities and paradoxes when faced with particular forms of knowledge (cf. especially “The Basic Paradox”). That is, these knowledge claims can only be rendered intelligible when the properties are attributed to things-in-themselves, rather than correlations between subjects and objects where we can only ever know the object as it is for-us, necessarily maintaining skepticism as to whether objects are this way in themselves.
I am thus beginning with the thesis that we have knowledge of certain primary qualities. Your question is a question of how we know these primary qualities. I think that’s a perfectly legitimate question, I just don’t have the answer to that question at this points.
Now before you jump all over me and ask “what warrants you in claiming that we have knowledge of certain primary qualities?”, let me respond by saying that my method is itself Kantian and transcendental (at least as I’ve construed Kant’s method). Like Kant with geometry, arithmetic, and Newtonian physics, I begin from the premise that certain scientific findings are true. I begin from the premise that we do have knowledge of the radioactive decay of isotopes in atoms, that we are able to calculate the speed of light and therefore how long light has been traveling to us from particular stars, and that neurology has shown us that mind and brain are one and the same things. My second premise is, as I argued in the Meillassoux posts, that these forms of knowledge can’t be rendered intelligible within the framework of the correlationist argument, but rather can only be rendered intelligible under a realist model of primary qualities.
As you have asked again and again, this obligates me to give some sort of account of how this knowledge is possible. We can share the premise that something is possible and does take place, without knowing how this is possible. For example, Darwin knew that evolution takes place through natural selection, but did not know how it took place, i.e., the mechanism of inheritance. It would take Mendel and later on Watson and Crick to identify the how of evolution. Something is similar here.
As an aside, it is possible that we might discover through inquiry that the kind of knowledge I’m hypothesizing is not possible, and that we only ever know of things in correlation. If that turns out to be the case, however, I believe we have to throw out the findings of these things pertaining to the ancestral and the neurological. I think that’s too high a price to pay.
Levi, I wouldn’t jump all over you, I am genuinely interested in the issue myself, I “challenged” you in good fun, of course…
Quid juris is a question of right, I’m not sure how it fits here, can you say a bit more about it?
This is as Kantian as one can be!
.
Would it be fair to rephrase it with a prescriptive as in “that we must have knowledge of primary qualities”?
So we have quid facti issue covered and not quid juris – I agree with you here actually, we can function just fine with knowledge we have without know how we have it – my issue is, of course, that I don’t think that without a criterion to distinguish between types of knowledge we can even have the basic “somehow I know it so we’ll start from here” kind of stance.
Hi Mikhail,
I see the question of knowledge of primary qualities as a quid juris question or a question of right, in that I am unable to see how we can talk about mind being brain or events prior to the existence of life without the premise that these are claims about things-in-themselves, not phenomena. That is, in order for these claims to be intelligible, it is necessary that these properties belong to the objects known and not simply be the result of a relation between us and that object (i.e., secondary properties like the taste of wine which are not in the wine, but a result of how the wine relates to our bodies).
I am not sure how this thesis is Kantian in character. It is transcendental in the sense that it refers to a condition for such knowledge, but differs markedly, I believe, from the Kantian thesis that we can only ever know phenomena or appearances and but must forever acknowledge that we can have no knowledge as to whether things-in-themselves are themselves this way. Where Kant, for example, claims that we cannot know whether or not things-in-themselves are spatial in the way we experience them, the realist would be committed to the claim that indeed they are spatial in this way and would be spatial in this way even if no human beings existed to experience them. So there is a transcendental dimension to my argument, but it is a “transcendental realism” rather than a “transcendental idealism“.
I would agree with the rephrasing of my thesis as you put it. It would run “if x is knowledge, then we must have knowledge of primary qualities.” I have mixed feelings about your final paragraph. I think scientists have all sorts of criteria as to how knowledge and truth can be distinguished, but usually these criteria are dissatisfying to the philosopher. They begin– depending on their object of study –from a rather naive realist stance, and cite criteria such as replicability in experiments, quantifiability, the absence of bias or intervening causal factors in experiments, consistency with well confirmed reigning theories, etc. These sorts of criteria get them through the day in their research, but clearly in the debate between the transcendental idealist and the transcendental realist these criteria are going to prove unsatisfactory as they don’t yet allow us to decide which epistemology is correct. The transcendental idealist in no way disputes the experimental method, but simply denies the realist interpretation of what this method discloses. The transcendental idealist will say “no harm is done in scientists believing that they are talking about the ‘real world’ (being naive realists thinking they are talking about things-as-they are in themselves), but their activity is only intelligible of we understand their findings to be about phenomena not things-in-themselves.” The transcendental realist sides with the scientist and says “no, the scientist really is discovering truths about things-as-they-in-themselves.” Regardless of how this debate plays out in philosophy, the scientist goes his merry way doing what he’s always done.
This might make this epistemology sound useless or without point, but the stakes are nonetheless high. As I remarked in one of our previous discussions, this particular epistemological debate isn’t, in my view, there to tell the scientist how to do his work, but is instead addressed to the skeptic. Both the transcendental idealist and the transcendental realist wish to refute the skeptic. They thus share the same aim. The transcendental idealist also wishes to curb dogmatism for all sorts of legitimate reasons. And, I think, the transcendental realist has a loathing for dogmatism as well. The transcendental idealist thinks that we can only render the universality of science intelligible if we begin from the premise that, to put it a bit crudely, objects conform to mind rather than mind to objects. The transcendental realist is bothered by the transcendental idealists solution because he thinks it allows for a new skepticism to sneak in the back door.
That is, if it is true, as the transcendental idealist argues, that we can never know objects-in-themselves but only phenomena or appearances, then the door is open to a skeptical critique of, say, evolutionary theory. In other words, the religious skeptic can always respond to the transcendental idealist by saying “you have argued that we can never know things-in-themselves but only phenomena. I concede that things appear to take place as evolutionary theory and astrophysics portrays them, but there is nothing to prevent me from holding the belief that things-in-themselves took place as the Biblical account of Genesis portrays them.” The transcendental realist thinks this is unacceptable.
As an aside, I don’t think the transcendental realist is committed to the claim that we can know everything about how things are in-themselves. For example, it might very well be true that there are genuine limits to our knowledge due to our inability to observe certain things or to devise technologies capable of measuring certain things. A good example of this would be quantum mechanics where we have all sorts of phenomena available to observation but where we really don’t know what these subatomic “entities” are in themselves. Here we might have a case where we have two competing theories of subatomic entities– let’s call them the Ptolemaic version and the Copernican version –where both theories are able to account for the phenomena observed, but where the intrinsic limits to our observational abilities prevent us from determining whether subatomic “entities” are “really” that way. Thems the breaks, if that’s the case.
Quick notes because I have to run (will read you guy’s stuff above).
Is it enough for the realist to say that the opposing view, that we don’t perceive real properties always either leads to the kind of incoherence that Schopenhauer pointed out vis a vis Kant- for example understanding causality to be only between appearances and also really in between things and themselves and appearances, or to what is in effect realism anyway, e.g. objects as thoughts in God’s mind?
As a positive answer I think Alva Noe’s Gibsonian/Heideggerian theory answer’s the question pretty well. Perception is primarily detecting opportunities for moving my body in relation to the environment, and the objects and properties we perceive are sets of such affordances. Of course a pigeon or bat will perceive different such affordances given their evolutionary niche.
But all these affordances are really there.
Anyhow, in a few hours I’ll go back and read what you guys have said above and post again.
So there are no issues in science regarding things like the observer’s interference in the observed phenomena or issues with methodology or issues with proof? No epistemological issues whatsoever? I find this rather surprising simplification/idealization of scientific discourse.
I think I mostly agree with your description of the issue (strange as it may sound as this point of our debate), and I do appreciate your honest admission that
a) we don’t know how to distinguish P-qualities from S-qualities, but we must (let’s call it Q-imperative) because otherwise we cannot claim that certain statements are intelligible;
b) there are limits to our knowledge, therefore that are some qualities of objects that we can never know.
I think one can easily reformulate your position in Kantian terms: the very distinction between P-qualities and S-qualities is basically a distinction between noumena and phenomena – we must posit this distinction (insert standard Kantian explanation of why). If we claim that certain qualities of objects that we are observing actually belong to the objects as they are in themselves, i.e. that certain S-qualities are in fact P-qualities, then we must have a criterion (insert my complaint how without a criterion, you project is not just weakened, it’s impossible) that would allow us to know the difference, not just posit the difference. In the absence of such criterion was must assume that we can only deal with S-qualities, period.
Now your argument from pragmatic use of common sense criteria (cannot demonstrate that bitterness of the wine is actually its P-quality, but it’s certainly its S-quality because I taste the bitterness) in science is attractive yet useless just like any common sense argument in our attempts to counter skepticism (if this is our agreed common stance). Let me think through some of these issues and respond in full. Again, I appreciate your willingness to engage these issues.
Mikhail,
I didn’t suggest that science shunts aside all epistemological questions– in fact I explicitly said that science has its own epistemology –I said that by comparison with philosophical epistemology its a rather crude epistemology. You are absolutely correct to point out that scientists are worried about the role the observer plays with respect to the observed. This is especially the case in psychology, ethnography, and quantum mechanics. The psychologist is deeply concerned with whether or not his presence biases the experiment in a particular way, whether through the assumptions he brings to bear on the phenomena he’s observing or the manner in which his presence biases his test subjects in ways that contaminate what is discovered. My point is simply that scientists are realists with respect to what they do. Moreover, their epistemological considerations are methodological considerations, not questions about whether or not what they discover is real or only “for-us” (the notable exception being quantum mechanics, but even then…).
I don’t think the Kantian solution you propose about p-properties and s-properties does the job. The transcendental realist will, of course, agree that we are the one drawing this distinction. The problem is that under your Kantian construal, p-properties are properties of phenomena, not things-in-themselves. For example, consider Kant’s famous example of the difference between the perceptual judgment that “the stone is warm” and the judgment “the sun warms the stone”. He tells us that the first judgment is a subjective judgment (a claim about s-qualities), while the second is an objective judgment (a claim about p-qualities). So far, so good. But Kant nonetheless qualifies that while we can know causal relations among phenomena, what is “for-us”, we have no idea whether or not the world itself has these causal relations. That’s where the transcendental realist and the transcendental empiricist part ways. Or rather, that’s the entire bone of contention. For the TR this causal relation would exist between stone and sun regardless of whether any consciousness ever existed to observe it, whereas for the TI this relation only obtains for humans with the very large restriction that we have no idea whether or not things themselves are like this. Thus, the TR will say “of course it is us how draw the distinction, but nonetheless this is a distinction about the referent in-itself, not the referent as it appears for-us.” So I guess I’m saying that I don’t see the debate as being a debate about whether or not we draw the distinction, but about whether or not these properties belong to things-themselves or whether we have to profess complete agnosticism as to whether or not p-properties belong to things themselves.
My offhand remark about wine was a bit stronger than you portray it. I was not suggesting that we cannot know whether wine in-itself has this particular taste, but rather saying we know that wine, independent of us, doesn’t have this particular taste property. We know this, for example, because this property of wine is variable depending on whether we are healthy or sick and because people’s palates differ. As such, we conclude that the taste of wine is a s-property because it only exists in the relation between us and the wine. By contrast, if mass is a p-property then this is because it remains the same despite changing relations to our body. Note I qualify this statement with an “if” because we now know that mass changes as a function of the speed at which an object is moving. This change in mass, however, is not an s-property because the change isn’t produced by a result of how our body relates to the object, but rather as a result of physical laws governing movement (i.e., the principles of relativity). Hopefully this clears up, a bit, as to just what I have in mind about the difference between p- and s-properties (which, incidentally, is a very nice abbreviation).
Just a quick thought, inspired by something Shahar said in a different context: Why, exactly, should we think that primary qualities belong to the thing in-itself and secondary qualities are products of a subjects relationship to a thing? Why isn’t the reverse equally plausible?
Now I know what the Cartesian answer to this question is; that doesn’t interest me, since it seems either to beg the question or not address it at all (primary qualities can be apprehended by the mind alone — OK, but that simply means that ‘primary quality’ can be thought independently of, or abstracted from a correlation, not that it refers to an entity that is ontologically independent of that correlation).
To fill the thought out a little, why would one think that the entities involved in ‘scientific explanation’ are more fundamental than correlational/phenomenological entities (and let’s face it, the real targets of QM are Husserl, Heidegger, and Gadamer)? — Especially since ‘scientific method’ operates with and theorizes the correlation between ‘subject’ and ‘the given’ (i.e. observation, abstraction, and experiment) in a manner that allows for certain forms of inference to non-apparent grounds. More simply put, without a firm grip on ‘givenness,’ you can’t infer a non-given ground. So why isn’t the grip on Givenness more ‘primordial’ than the the non-given that we infer from it? Why isn’t the fossil more fundamental than what we can infer form it?
Alexei,
I’m actually sympathetic to some of what you say here. I think one of the mistakes of some versions of realism is to treat secondary qualities as a synonym for “unreal”. So primary qualities are said to be real and secondary qualities are treated, as it were, as illusions. If realism denies ontological dualism or the thesis that mind and body are distinct substances, then secondary properties are just as real as any other properties that exist in the universe. The difference is that they are dependent in their existence, requiring a set of conditions to obtain in order to occur. The gripe on the part of the realist lies in the anthropocentricism of the anti-realist. Because the anti-realist essentially treats all properties as s-properties or dependent on humans in some way or another, the anti-realist is committed to the thesis that the notion of a universe independent of humans or without humans is meaningless where knowledge is concerned.
It seems to me that when you talk about the priority of s-properties over p-properties you’re conflating an epistemological issue with an ontological claim. The realist certainly agrees with the anti-realist that the fossil must be given or available in order that inferences be made from it. In other words, the realist is not making the claim that we can somehow know independent of engagement with the world (at least, I’m not). But this is different than the idea of “correlation”. Yes, I must relate to the fossil to make inferences from the fossil. But for the realist, the fossil is not dependent on my relating to it. That is, it has properties that are not “s-properties” and possesses these properties regardless of whether I relate to it. The realist is even willing to agree that there are cases where no inferences can reliably be made from the given, as in the case of quantum mechanics where we really don’t know what these quanta are in-themselves but only know what they are for-us.
This point can be illustrated further in terms of neurology. The realist neurologist agrees that we need good phenomenology to do neurology. That is, the neurologist rejects the thesis that we can simply do fMRIs or look at neurons under a microscope and infer what these neurons or blood flows are doing. This is because these activities, severed from phenomenology or what the neurologists call “cognitive science”, are an explanans without an explanandum. Sure, we observe all sorts of interesting blood flow, electrical processes, and chemical processes taking place when we use fMRI or observe a neuron under an atomic microscope, but we have no idea what the interactions we’re observing are explaining or doing. Nonetheless, the cognitive psychologist and the neurologist differ from the phenomenologist in two crucial respects. First, whereas the [Husserlian] phenomenologist takes these descriptions of intentional lived experience as being apodictic (here Merleau-Ponty would be exempt from this criticism as he was willing to include the findings of empirical psychology in his phenomenology), the cognitive psychologist holds both that these descriptions are often mistaken and can be shown to be such through careful observation and experimentation, and second, the neurologist holds that the neurological discoveries can often show that phenomena at the phenomenological level are, in fact, illusory (the distinction between folk-psychology and scientific psychology).
Part of the problem with philosophical phenomenology as practiced in Continental philosophy (and here again Merleau-Ponty is exempt) is that it bases its findings on normal functioning. Because it ignores brain-damaged patients as well as those suffering from various psychic maladies, it ends up with a distorted understanding of just how the mind works, so its “given” is highly contentious where its “apodicticity” is concerned. Consequently, a major difference between the cognitive psychologist and the (non-Merleau-Pontyian) phenomenologist, is that the latter’s methodology is foundationalist, while the formers methodology is dialectical (very loosely construed). That is, the phenomenologist begins with the premise that his description of experience or the given and givenness is a foundation to which everything must be traced back and which cannot itself ever be overturned by subsequent findings of an empirical sort. For the cognitive psychologist, by contrast, it is agreed that we have to have good descriptions of what mental life is, but (and here’s the dialectical moment), these descriptions get revised and significantly modified as a result of our experimental observations (at the cognitive or phenomenological level), as a result of our observations of brain damaged patients, and as a result of what we discover about the neurological structure of the brain in relation to these conscious level phenomena.
Additionally, where the classical phenomenologist will deny that brain could be a condition for consciousness (Husserl’s notorious claim that nature cannot be a condition for consciousness and that the entire world would remain as it is for the transcendental ego even if nature or the physical world were destroyed), the cognitive psychologist and the neurologist will begin from the premise that brain is a condition for consciousness. In short, the neurologist cannot concede the Kantian thesis that brain is only what it is for-us or that it is only a phenomena because brain is a condition for phenomena. The neurologist will agree that there are many things we might not be able to know about brain. For example, while we know a lot about neurons, we as yet have very little knowledge about networks of neurons, how they are formed, and how to individuate them. But this claim is entirely different from the claim that we can only know phenomena.
Before we get down to it, Levi, might I suggest that we leave aside various positions that we disagree with, and simply discuss the problem at hand? I ask only because I really don’t want our conversation to turn into something like “this is how I read Husserl and Merleau-Ponty,” and I’m already feeling a little fidgety with your presentation of Husserl (you seem to have flattened out the difference between Evidence and lived experience, for instance), who really isn’t central to the point at hand. Now I know you’re bringing him and Merleau-Ponty up in order to make your points more concrete, but this concreteness depends so much on a particular interpretation that we risk getting sidetracked by them and how best to understand what they’re up to.
So with this said, let’s move on. The thrust of my question wasn’t about the ‘reality’ of secondary qualities to primary ones. There’s a potential problem there only to the extent that one wishes to assert some form of type-type identity between them (in which case you’re committed to some form of eliminative realism). the thrust of my question was actually about how one assigns priority to competing explanations, which is probably just a variation on Mikhail’s question concerning how one knows that there are primary qualities. That is, a causal explanation, which QM insists is ontologically primary, only gets us so far and ultimately isn’t very informative. Moreover, a causal explanation has antecedent conditions for its intelligbility. Call this latter one something like logical explanation (certain epistemological conditions make causal explanations possible). One can’t begin to offer a causal explanation of something, unless one has a sufficiently robust conceptual framework to couch the explanation. These competing orders of explanation (casual and logical) are usually sorted out historically (paradigm shifts, or whatever), but it doesn’t change the overarching point: whether it is realist or anti-realist, every theory is comprised of at least two orders of explanation. What I would add is that to emphasize one over the other is to distort matters.
Furthermore, I don’t think it’s helpful to appeal to what a scientist thinks she is doing when she is working, since she is just as proe to confablation as any other psychological test-subject. whether the Scientist-herself is a realist matters very little. What does matter is how neatly epistemological and ontological claims can actually be separated while maintaining a claim to knowledge. For something to be a candidate for ‘knowledge,’ it needs to satisfy three criteria: (1) it must be explanatory, (2) it must be predicitive, and (3) it must be decidable (capable of being either true or false and of being demonstrated as being one or the other. Now, take Graham Harman’s argument for infinitely withdrawn objects as an example. It’s a nice argument, but it will never amount to knowledge, since he his model has no predictive power, I have no idea whether it’s decidable, and I’m not sure what it explains. I’m being hard on Harman, but he can take it, I’m sure. My point, however, remains: If Harman is a realist, then there’s not much point in siding with him, since his account privileges a logical order of explanation and excises any kind of causal explanation (despite being very concerned with the concept of ‘causality’).
Now, a couple of more concrete points: about this:
I don’t think so, since I asked a question, I didn’t assert anything. There happen to be at least two orders of explanation operant in any theory (which I’ve called logical and ontological), I don’t see how asking after their priority, and trying to point out that the decision concerning priority is not itself ontological, conflates the orders of expalanation at all. again, the issue of priority needs to be sorted out. Why should I think that ontological/causal expalantions are prior to logical epistemological ones, without begging thereby demonstrating that the logical is prior to the causal — at least at the level of explanation?
No, they are the qualities of things – they are not known to us, that’s Kantian twist. Actually, as far as Kant is concerned, p-qualities and s-qualities could be the same thing, we just don’t have the ability to demonstrate it, because we don’t have a criterion to distinguish between the two – as Alexei put it some time ago, we cannot get out of our skin.
P-property do belong to things themselves, it’s how they get to be called p-properties, it’s in their definition – we posit them as properties that belong to things themselves, we just cannot experience them as such and know what they are.
I don’t really understand then what you mean by a phrase “we draw the distinction” – the distinction is posited, it does not come from experience, i.e. it’s not as if you notice the difference between p-qualities and s-qualities. In TR, you say, you accept that it is we who draw the distinction, but if it is a theoretical distinction that you simply decide to impose on a variety of available qualities, without a proper criterion you cannot even draw the distinction.
For example, if I said something like this: “There are two types of people in the world : smart people and stupid people.” Yet I provided no criterion of what constitutes smartness, I am unable to respond to a very basic inquiry like: “How do you tell which ones are which?” In other words, you can of course claim that there are p-qualities and s-qualities, yet the distinction is rather baseless without a criterion.
I can of course come up with an empirical criterion like: “People who wear glasses are smart, people who wear they hats backwards and pop the collar of their polo shirts are stupid” and it’s going to work in many many cases, but it’s hardly scientific and, what is most important for Kant, at least, necessary and precise.
I get your wine example, I think, and in this case you’re saying something like: “We can try to formulate the criterion of the distinction between p-qualities and s-qualities in terms of constancy/change” yet you quickly add that it’s not a very solid criterion because things like mass change etc etc. In this case are you disagreeing with QM in his assertion that p-qualities are mathematizable qualities, because mass is clearly mathematizable?
Again, I know I keep repeating this point, but you keep phrasing things in a way that makes me nervous: treating all properties AS s-properties is not the same as denying that p-properties EXIST “regardless of whether I relate to it” – as I stated before, it is necessary for Kant to argue that thing themselves possess p-properties (although it awkward to phrase it this way), i.e. we must have things-in-themselves in order to have then things as they appear to us. I know we’ve already covered this, but I’d like to make sure we distinguish between a) stating there is a difference between p- and s-qualities, and b) stating that we can know what that difference is. You said that we cannot yet demonstrate how b) is possible, so you are still in Kant’s camp for now.
I want to comment on this too, now that I think about it,
Depending upon what exactly you mean by this, I might be inclined to agree and affirm it. Here’s why (1) knowledge isn’t independent of knowers. (2) meaning isn’t independent of consciousness. So, to the extent that you need humans in order to have a universe with meaning and knowledge in it, knowledge and meaning depend upon humans.
Put slightly differently, the laws of physics aren’t ‘out there in the universe,’ there ‘in here with us humans.’ ‘Out there’ = bare individuals that interact. ‘In here’ = semantically rich, lawful articulations of interaction types. Absent an ‘in here,’ there’s still an ‘out there.’ It just doesn’t have any laws, meaning, or knowledge.
Hi Alexei,
I don’t think we’re in disagreement as to what’s involved in explanation. In fact, you’re repeating a number of points back to me that I made earlier in the thread pertaining to precisely this point and in my initial response to you (the bit about explanans and explandum). I’m fairly in agreement with you as regards Harman’s realism. Far from solving the correlationist problem by showing that all objects withdraw from one another (and not just in the human-object relation), it seems to me that he exacerbates the problem. I am less keen on your second criteria where you assert that all knowledge must be predictive in character. For example, a good deal of chaos and complexity science isn’t predictive in character but rather explanatory. It may be that this just isn’t knowledge, but I don’t want to jump to that conclusion so quickly.
I have also repeatedly emphasized that the scientist’s self-understanding doesn’t decide this epistemological issue. The scientist is a realist, but none of this decides whether or not realism is true.
In my initial post responding to you, I was not asserting that ontological ontological and causal explanations are prior to logical epistemological explanations. I was reminding you that we need to recognize that epistemological claims and ontological claims are two distinct kinds of claims. I have no disagreement with your thesis that something must be “given” (broadly construed) for us to set about explaining it. That’s never, as far as I’m concerned, been in dispute and even goes without saying.
My point, is that we shouldn’t confuse this priority of how explanation works with whether or not certain properties in question belong to the things-themselves or not. And it is here where you and I, I think, have a disagreement. Approaching the issue from a Kantian angle, you hold that we can only ever know phenomena and can never know whether or not things-in-themselves themselves have these properties. Is that right? For example, Kant tells us that we can never know whether or not entities are spatial in the sense that we experience them (in his stronger statements he even says that they are not spatial, cf. http://larvalsubjects.wordpress.com/2009/04/06/time-turtles-kant-and-correlationism/). The TR necessarily rejects this thesis as he believes that it would require us to reject wide swaths of our hard won knowledge because these knowledges require us to make claims that exceed the given or claims about things-in-themselves. The TR, for example, does not think the counter-factual solution (had a subject been there to observe x it would have been like this) is sufficient for talking about a world prior to life.
When the realist argues that such and such a property belongs to humans whether we exist or not, he is making an ontological claim rather than an epistemological claim. His rejection of the anthropocentrism of the correlationist position isn’t a rejection of the facile or trite observation that “we had to discover x, so it is related to us”. He has no objection to this observation. What he objects to is the thesis that we only have knowledge of phenomena (the “for-us”) and never things-in-themselves. In other words, the correlationist is making a much stronger claim: viz., the claim that all objects of knowledge are only appearances for us, without the ability to know whether the world independent of us has these characteristics.
Mikhail,
I understand your point better now. As far as I’m concerned, p-properties lose all interest if we can’t understand them. For the realist (or this realist) if they are only of interest if they can somehow be known. In the absence of being able to know these p-properties, Kantianism (or some other variant of correlationism) is the only variant. We are modest realists if we take this route in the sense that we concede, as the philosopher-physicist d’Espagnat puts it, that there is a “something” beyond our experience of the world, but beyond this minimal knowledge we can know nothing of this something. This is why I was perhaps a bit overhasty in adopting the term “object-oriented philosophy” for my realism. As a result, everyone took me to be adopting Harman’s metaphysics, but I don’t see how Harman’s metaphysics escapes this sort of problem. Anyway, to clarify, p-properties are only of interest if they can be known.
I am not convinced that the distinction between p-properties and s-properties is posited rather than the result of observation. Then again, I suppose this depends on what you mean by “posited”. I say this because it seems to me that we arrive at this distinction by virtue of a number of observations about objects where we gradually discover that certain properties aren’t in objects but only in our relation to objects. The wine is an example. I thought the wine was sweet. I get sick or eat a good cheese before sipping my wine. I discover that the taste of the wine changes. I conclude that the taste of the wine must have something to do with me rather than the wine itself. By contrast, every time I take the ph of the wine, it comes up the same. I therefore conclude that ph is in the wine.
With respect to Meillassoux, I have not yet decided whether I share his thesis that p-qualities are those qualities that are mathematizable. I think here a lot of work needs to be done to explain just what “mathematizable” means. I mean, clearly I can count the United States and Russia as two countries. Is this an example of mathematization? Are the United States and Russia entities in the robust metaphysical sense? It doesn’t seem to me that they are, but rather seems that the U.S. and Russia are s-properties in the sense that they depend on the relations of their people and the recognition of other countries to exist. In other words, I don’t think we can just wave our hands and say “if its mathematizable then it’s a p-property.” We need something more precise than that, or a specification of when and under what conditions m-properties (mathematical properties) are p-properties.
I don’t think this is a small issue in Meillassoux’s thought. I very much admire the first and second chapters of After Finitude, but I think QM gets himself in trouble after that. Because he has begun with the thesis that p-properties are m-properties, he thinks that he’s authorized to reason from abstract mathematical principles to properties of being. Thus, for example, he concludes that everything about being is characterized by contingency. Yet we all know there’s a distinction between mathematical or logical necessity, and natural necessity, which he’s entirely ignoring. Gabriel Catren has really taken him to task over this. At any rate, my point about mass wasn’t that it isn’t a p-property, only that it is variable under conditions of speed. That variability, however, is mathematical in nature, i.e., we can predict it and know exactly how it will change given such and such a speed.
One final point, I am not sure I share the thesis that we cannot get out of our skin. It seems to me that in our sciences we use all sorts of technologies (observational equipments) and mathematics to “get out of our skin” or escape the confines of our own crude observational apparatus (our five sense). Thus, for example, we devise measuring devices for forms of light that are above and below the spectrum of light we can sense through our own five senses. Likewise, we build detection devices deep in the earth filled with very pure water to detect neutrinos (which can pass right through the earth in most circumstances). We build the Haldron super-collidor to find gravitrons, etc. Einstein and the quantum physicists had to devise forms of mathematics to generate concepts no longer grounded in the immediacies of the “simple ideas” described by Galileo and Descartes (position, weight, length, duration, etc), generating concepts that exceed those of phenomenological lived experience. In other words, I do think there’s a sense in which we “get out of our skin”.
In reading this line of dispute I cannot help but think how the Analytic Branch of philosophy framed the Realist vs. Anti-Realist debate in a much more elucidating manner. On the Analytic side the question comes right back down to Mikhail’s original question. The question is of justification. Analytical Realists think that one can justify one’s knowledge of things by virtue of how they really are. For Anti-Realists one can only justify them by appealing to reasons (which are either communially constructed, or internal in a Kantian sense, categories which serve as both reasons and causes). Because Mikhail frames the question as “how do you know this” one is either forced to appeal to reasons why you know, or some causal mechanism which leads to this knowledge (Kant conflates the two).
Levi’s position: “My thesis is that we require primary qualities to belong to things-in-themselves for certain knowledge claims to be intelligible. That is, knowledge of primary qualities is a quid juris requirement for certain knowledge claims.”
strikes me as a confusion of questions of justification with questions of causation, something which ends up being a semantic dispute, what is “knowing”? Someone like Wittgenstein who wants to straighten out what is a cause and what is a reason, claims that the kind of “knowledge” that Levi is speaking of really shouldn’t be called knowing, since in a regard it cannot be wrong. As he frames it, part of the grammar of “to know” is the capacity to “be wrong” about what you know. The primary quality that Levi arches toward simply isn’t knowledge as we usually think of it, something we can be mistaken about. It is simply, as he says, “a requirement”. The dispute strikes me as a confusion of uses of the word “know”.
What I suspect the answer is, or at least a very good way of dividing it up (if you are not going to allow a Kantian conflation of reasons and causes), is that indeed Levi’s Realist need for things really to be what they are outside of our thoughts about them is the “causal” source of our knowledge, while our historically constructed criteria which is shared between persons is the source of our justification of our knowledge, such that the two are interbraided.
Yes, I think we’ve now identified the precise form of the problem:
That strikes me as right. This said, I don’t think you’ve accurately formulated the alternatives, precisely because this particular problem is not specifically Kantian (although Kant does propose a solution). The general problem, the logical one, the one with real bite, seems to be this: ontological commitments are only coherent — rational — in virtue of explanatory commitments (Occam’s razor is an instance of this; so is Quine’s maxim, “To be, is to be the value of a bound variable’). The converse (epistemological commitments are coherent only in virtue of ontological ones) doesn’t seem to be compelling; in fact, it strikes me as what creationists want to say.
As for the whole transcendental realism thing, I must confess that i’m totally unconvinced. I just don’t understand the motivation behind trying to claim that we can know things in-themselves, since the very Idea of a thing in-itself is supposed to show that empirical knowledge is exhaustive of knowledge in general — anything left over, like Harman’s withdrawn objects, simply isn’t a candidate for knowledge claims. What else do you want, really? I fail to see what large swath of hard won knowledge anti-realism abrogates, by saying that things in-themselves, infinitely withdrawn objects, aren’t candidates for knowledge claims; we very well may be able to think them, to speculate about them, but if they do not figure in theories that allow for explanatory, predictive and decidable results, then they’re simply not knowledge.
So forget things in-themselves for a moment. Think of String Theory. Lee Smolin has argued that string theory has been a disaster for theoretical physics precisely because it has no predictive power, and hence no decidability. It doesn’t count as knowledge, then, not because it describes some unobservable, pre-givenness, but simply because there is no way for us to ever answer the question ‘how do we know that string theory is serviceably true?’ String theory is explanatory, sure, but (to be hyperbolic) it’s only marginally better than creationism — for neither are predictive or decidable.
Here’s another nitpicky example: We use chaos theory to predict weather patterns and whatnot (one of the reasons we don’t have reliable projections of weather patterns past 7 days, has to do with computational limitations). So chaos theory is a theory. And most systems theoretical accounts have a fair bit of predictive power, once they’re actually worked out. The problem, however, is that complexity theory (for instance) is little more than a series of grand statements and gestures; folks haven’t worked out the mathy bits yet. So it has explanatory power (maybe), but isn’t decidable, or predictive.
All this said, what isn’t a theory right now, might very well be worked out in the future. But to base one’s ontological commitments on a future that may never come is kinda weird.
Alexei,
Your remarks here are interesting. First, I would agree with you concerning string theory, so that’s neither here nor there (you’re mistaken, btw, in the claim that it doesn’t make predictions. Rather, the problem is that it doesn’t yet give us any predictions we can test). Until string theory can provide some sort of experimental verification it remains “just a theory”. No problem here with that thesis.
What I find interesting in your remarks is the charge of “creationism”. I find this interesting because I myself have charged correlationism with being equivalent to creationism. This must be the “catch-all” insult in philosophy these days. I say that correlationism is equivalent to creationism because, in claiming that we only ever have knowledge of phenomena and cannot know whether the properties we discover belong to things-in-themselves, the door always remains open for the obscurantist to rejoin by saying “I quite agree that things appear to be the way the scientist describes them, but the TI has shown us that we can never know things-in-themselves. Thus, while things certainly appear to be far older than life, at the level of things-in-themselves the Biblical account of Genesis is true and they are really only 6000 years old.” I don’t think it’s a mistake that we’ve recently witnessed a theological turn in postmodern and phenomenological philosophy. In both cases, I think this move has been rendered possible by the correlationist loophole.
You write:
This is a very nice formulation of what Bhaskar calls the “epistemological fallacy” or the thesis that ontological claims can be reduced to epistemological claims. The realist thinks that this is a fallacy because he endorses the thesis that the being of something isn’t reducible to the manner in which we know that thing. The realist (or, at least me) is more than happy to say we have to discover things and that we can be mistaken about these things– here I think kvond misses the point –but is also committed to the thesis that what we discover is real in the robust sense (not merely as phenomena). I am not sure what you mean in your statement that “epistemological commitments are only coherent in virtue of ontological ones.”
Kvond,
I think you’re missing an important nuance in my position. I have nowhere suggested that it is impossible for us to be wrong. We are wrong all the time and we put forward many theories that turn out to be mistaken. My posts and comments have been littered with examples of such mistaken hypotheses such as the caloric theory of heat. My thesis is that if something is knowledge then it refers to something real, not something that is simply a phenomena or appearance for us. I suspect that you are arriving at this characterization of my position based on a confusion about the argument from ancestrality. I begin from the premise that evolution did take place and there was a time prior to life. That is, I begin from the premise that the theory of evolution is knowledge and that we are in consensus that this is knowledge. Now, if this is knowledge, it leads to the conclusion, I believe, that realist theories of knowledge must be true because claims about things independent of humans are meaningless within the constraints of anti-realist theories of knowledge. The question then becomes your question of how exactly we have this knowledge. We know that we have this knowledge, and we know that this knowledge refers to something in-itself (in my view anyway), so how is this possible? Note that this line of argument could be entirely overturned tomorrow if rabbit bones were found in the pre-Cambrian strata of the fossil record. That is, it’s based on evolution being true. I’m not holding my breath for such a discovery, though.
LS: “The realist (or, at least me) is more than happy to say we have to discover things and that we can be mistaken about these things– here I think kvond misses the point –but is also committed to the thesis that what we discover is real in the robust sense (not merely as phenomena).”
Kvond: I am interested in this wrongness. Perhaps give an example of a primary quality that were were wrong about, a quality that cannot already be qualified as a secondary quality.
Seems we were posting together.
LS: “We know that we have this knowledge, and we know that this knowledge refers to something in-itself (in my view anyway), so how is this possible? Note that this line of argument could be entirely overturned tomorrow if rabbit bones were found in the pre-Cambrian strata of the fossil record.”
Kvond: In my view when you say “refers to” there is division. When one attempts to justify the claim, one refers to shared criteria (not the thing itself). But it also “refers to” the causal relationship these criteria are assumed to have (must have due to their rational coherence), to some feature of the world.
When you were wrong about pre-Cambrian rabbits, you were wrong in your use of criteria. The same world caused the two positions.
I think that for this reason it is much better to talk about “beliefs”.
Take an example: Let us say we are looking at a chessboard like the one Mikhail puts the top of the page. We want to say, “The King is on square E1” is a true sentence. Someone else says, how do you know? Well, “This is the E column, and this is the 1 row” and you agree. Then you move your finger and see that you were somehow wrong, the King is on D1. Now, were you wrong about the primary quality of the King?
One still assumes that something about the King (and its place in the world) some Real Thing is causing your beliefs and guiding your application of knowledge, but the truth or falsity of your sentence relies upon the justification of agreed upon claims. At least as far as I can see.
Perhaps though you have a different example in mind.
Kvond,
It sounds like you’re well advanced of where I’m at on this issue. I’m at the point where I believe I can say that we have knowledge of real things in the world, but not at the point where I can say how this knowledge is possible.
An example of being mistaken about primary qualities would be that of the caloric theory of heat. One of the running theories was that not substances had a substance within them called “caloric” whereas cold substances did not. Certain predictions followed from this: hot objects would have a greater mass than cold objects, and hot objects could not remain hot indefinitely (because they would lose their caloric over time). Both of these predictions turned out to be mistaken. In the first case, there was no difference in mass between hot and cold objects. In the second case, the boring of holes in iron to make cannons showed that objects could remain hot indefinitely. For a while the advocates of the caloric theory tried to argue that caloric was a strange substance without mass, but they weren’t able to come up with any experimental scenarios to test this hypothesis. When the kinetic theory of heat came along (that heat is agitation of molecules or particles) it was able to account for a broad range of phenomena while also shedding light on other phenomena that, on the surface, appeared unrelated to the phenomenon of heat. This would be an example of a claim about primary qualities that was mistaken. I guess, then, I’m saying that science is very much a question of discovering what the primary qualities are. It gives many hypotheses as to what they might be with respect to a particular phenomenon and a number of these hypotheses turn out to be false.
Your example of chess doesn’t quite work for me, because I don’t think chess and chess pieces can be thought in terms of primary qualities. “Being-a-queen” is a property that belongs to a chess piece only in relation to us. Proof of this lies in the fact that should we lose the queen belonging to our chess set (not in the game) we can replace it with a quarter and stipulate that the quarter is the queen. In other words, it’s nothing about the object that makes it a queen, nor anything about the board that makes it a chess set. It is only in relation to a set of collective relations that something becomes a queen or a game of chess.
LS: “An example of being mistaken about primary qualities would be that of the caloric theory of heat. One of the running theories was that not substances had a substance within them called “caloric” whereas cold substances did not.”
Kvond: This seems to be a good example. I’m glad you provided it.
I’m try to parse a “theory about heat” from knowledge of the primary qualities of certain states in the world. Now, if I’m following you correctly, when I say that this glowing hot piece of iron has some “caloric substance” within it, you are claiming that this is a mistake about the primary quality of the iron. But why would this caloric substance quality be primary? Would not the iron ingot still be an iron ingot with or without the caloric substance?
As far as the chess example, the spatial location of the King is not really equivalent to “being a King” which is purely defintional. One presumes that the King “really” is on one square or another, and that this state is the cause of the truth of the sentence. Whether the King really is a King is not of this order.
As I understand it, the caloric theory of heat wouldn’t be a statement about iron, so much as a statement about caloric. That is, it asserted the thesis that there’s a substance that exists in its own right known as “caloric”. Later on they found that heat was just an agitation of, as per your example, the atoms that make up iron rather than an independent substance in its own right.
I’m agreed with your point about the position of the King and thing I was making a similar point less eloquently. The realist point would be that where something like a King can only be a King by virtue of this definitional determination, the properties of the elements on the Periodic Chart belong to those elements regardless of whether we define them. That is, the definition doesn’t make hydrogen “hydrogen”, but only reports the properties that already belong to hydrogen.
LS: “the properties of the elements on the Periodic Chart belong to those elements regardless of whether we define them. That is, the definition doesn’t make hydrogen “hydrogen”, but only reports the properties that already belong to hydrogen.
Kvond: To give my two cents, I suppose I am too much of an Anti-Realist (really a Non-Realist) in terms of justication to find this presumption covincing, or entirely coherent. Sentences made about hydrogen that are determined to be true are historically made by animals whose behaviors are contingently oriented. I prefer to talk of sentences as existing at a time and place in history, and therefore are contingently true, and and not true for all eternity. The ascription of the properties of hydrogen are ever provisional, and consitute knowledge only within our communial space of shared criteria. When we say “hydrogen has property x” this is something we are doing, it is an act, and therefore relies upon the rest of the social field for its reference and importance. The idea that the periodic table exists “out there” is for me the equivalent to saying that the Universe speaks English.
In my mind the entire desire for there to be some facts of the world which make a certain number of our statments true for all eternity is simply a vestiage of the old feeling that we need some supra-human authority to make our sentences (and beliefs) true, to guarantee, to gold stamp them. In the past this was called “God”. Now it is is called “Reality”.
I do though understand better your claim. I appreciate the clarification.
This is what I find the most objectionable part of this “modest realism” – you keep saying it and I keep objecting to it – to concede that there is a something beyond our experience is not yet realism, Kant does the same thing and he is not, according to you, a realist. Your definition of realism should contain some sort of an explanation about this “minimal knowledge” otherwise it’s just an unjustified assumption without any demonstrable proof, a dogmatic statement not unlike a religious pronouncement or a mystical “intellectual intuition”…
Agreed. But since we cannot explain how they can be known, and simply stating that somehow they are cannot constitute serious philosophical position, we should just throw away p-qualities.
I don’t know about that, I find this to be an interesting question personally. My tendency is to say that with scientific instruments we haven’t really changed much of our perception of reality, just a quality of data that can perceive, but the jury is still out on that one for me.
Mikhail,
I don’t understand why my observation about Kant is akin to religious conviction or an unjustifiable thesis. I’ve never suggested that Kant is a Berkeleyian idealist who holds that “esse est percepi”, only that we don’t get a whole lot from the thesis that there is a thing-in-itself without ever knowing the thing-in-itself. It seems to me that these debates over realism and anti-realism are ultimately debates over truth. Now were I to grade anti-realists from best to worst, I would place figures like Kant and Husserl towards the top of the list as best and figures like Lyotard, Wittgenstein, and Baudrillard among the worst. At least in the case of Kant and Husserl we get an attempt to account for something like universal knowledge, even if it is only ever restricted to phenomena. When we get far down the correlationist spectrum to folks like Lyotard, late Wittgenstein, and Baudrillard, all we get are competing “world-versions” that are unable to claim any superiority over one another. At any rate, the issue I think boils down to what we believe the status of knowledge to be. Is it about the real world or is it only about appearances.
I am perplexed by your thesis that if I am unable to yet give an account of how knowledge of p-properties is possible, I should throw out p-properties altogether. I think I have given compelling arguments that we have this knowledge. The fact that we have this knowledge thus presents us with a philosophical problem to be solved. Suppose I’m wrong and Kant (or some other correlationist is right). For centuries people did geometry and arithmetic before Kant. Newton died three years after Kant was born. Are you really suggesting that the mathematicians and Newton had to await Kant before they could engage in their researches? That’s an odd thesis.
Your final statement about scientific instruments and changes in our perception of reality is ambiguous to me. If you are making the claim that our scientific instruments have not literally changed how we perceive the world (i.e., the phenomenology of our experience) then I agree. We still experience the world in terms of profiles, foreground and background, nearness and distance, and all the rest. If you’re using the term “perception” in a more figurative sense to mean “we still don’t conceive the world differently, then I strongly disagree. I think the discoveries of science have deeply transformed our understanding of the world even for the layman who only has a very limited knowledge of science.
Kvond,
I appreciate your remarks but find them very strange coming from a Spinozist. I, of course, agree that statements about the world are historical in the sense that they’re discovered at a particular point in time. I also agree that we can subsequently discover that we were mistaken in our understanding of these phenomena. I am not willing to go all the way with Latour and say that claiming the Black Plague in Europe during the Middle Ages was caused by the Yersinia pestis bacteria is an anachronism because bacteria only exist follow Pasteur. Regardless of whatever the folks of the Middle Ages might have thought, I am committed to the thesis that the Bubonic Plague was caused by this bacteria and that this isn’t simply a historical interpretation or result of a “definition”.
As a sidenote:
LS: When we get far down the correlationist spectrum to folks like Lyotard, late Wittgenstein, and Baudrillard, all we get are competing “world-versions” that are unable to claim any superiority over one another.
Kvond: With this I would have to disagree, at least to say that late Wittgenstein had great influence on Donald Davidson who uses his concepts of the impossibility of Private Language to propose a definite Theory of Truth, and a theory of rational ascription which makes discourse universal (denying that there is any sense of speaking of “conceptual schemes”).
While Davidson’s universal translatability certainly does not allow such a thing as “universal knowledge” I don’t really know what “universal knowledge” gets you other than an claim that what I know is true for all time and eternity. This is distinctly in my view a Religious inspired claim.
I wonder where one places Latour in this game. He certainly seems to be on the “wrong” side of the universal knowledge requirement. Have you lessened in your attraction to Latour?
Really fascinating discussion here. I wish I had the energy to contribute to it more! I’m in broad agreement with Levi, but I think Mikhail, Alexei and Kevin are bringing up great points. Just two quick comments on Mikhail’s last comment:
“This is what I find the most objectionable part of this “modest realism” – you keep saying it and I keep objecting to it – to concede that there is a something beyond our experience is not yet realism, Kant does the same thing and he is not, according to you, a realist.”
Let me know if you’ve already answered this (it’s difficult to even keep up with all these discussions!), but isn’t Meillassoux’s main point in AF that this form of weak correlationism is untenable – requiring one to adopt either absolute correlationism, or Meillassoux’s own position, which entails knowledge of the absolute? That is to say, hasn’t Meillassoux already shown that logically, the weak correlationist already has knowledge of the absolute?
And my second point,
“My tendency is to say that with scientific instruments we haven’t really changed much of our perception of reality, just a quality of data that can perceive, but the jury is still out on that one for me.”
I think this is a debatable point as well, but off-hand I can think of two pieces of evidence that science has actually changed our perception – one, the bizarre mathematics involved in cutting-edge physics, which model unvisualizable, multi-dimensional space-times. And two, psychological experiments showing that unperceived visuals (e.g. a word flashed too quickly for one’s consciousness to pick it up) are real and have effects. Both of these examples are, in principle, impossible for us to ever experience, and are only possible with the aid of scientific instruments.
Allow me a simple illustration:
Imagine that you are looking at an object, any object, and you list all the properties that it has – let’s say you can only come up with 10 properties. Now because you have a basic distinction (whichever way you came up with it) between p-qualities and s-qualities, I ask you: “Out of these 10 qualities, which ones are primary and which ones are secondary?” You say: “I don’t know, but I do know that some must be primary and some must be secondary.” Is this what you’re saying at this point of our discussion? It seems that this is what you are calling “modest realism,” that is, you tell me that some of the 10 qualities must be primary, we just don’t know which ones to which I can only reply something like – “Let me know when you do know because as far as we are concerned now, all of these properties could be primary or all of them can be secondary, or maybe some are even tertiary”.
Ok, so you seem to be saying something like this in return: “Hold on, I think I can sort of intuit that there is a difference between properties that are sort of subjective and other properties that seem to be more or less permanent or stable.” To which I can only say, again, with all due respect: “This does not sound like a good philosophical theory, but more like an intuition or a hunch.”
If we are indeed talking about a hunch or an intuition, how can we even raise the question of truth?
LS: “I appreciate your remarks but find them very strange coming from a Spinozist.”
Kvond: Actually, there are very good arguments that Spinoza presents a Coherence Theory of Truth (statements are made true via their mutual coherence), rather than a Correspondence Theory of Truth. And, as I have pointed out to you before, I do not believe that one could assert that human beings can hold completely adequate ideas under the threshold of their defition that Spinoza provides. Not even the propositions and definitions in the Ethics are eternal truths. Only truths as mutally coherent as Spinoza could make them. As a Spinozist of a kind, what I most appreciate is the way that he directs our vision towards the causes of things. But our knowledge is always historically bound.
LS: “I am committed to the thesis that the Bubonic Plague was caused by this bacteria and that this isn’t simply a historical interpretation or result of a “definition”.
Kvond: I am committed to this thesis as well, but as a thesis. And as a thesis it depends on definitions, and therefore in historically contingent in its truth. The “defintions” that this thesis depend upon are not made up in the middle of the air, but rather come from active engagment with the world, and the demands of pragmatic benefits of coherence. There is no extra secret sauce of correspondent knowledge that makes the thesis true. At least in my opinion.
Levi, what is the difference in practice that is produced if your theory is true? What changes? What is one able to say or do?
Kvond,
I think that one notable difference created in practice is that we, as philosophers, will take empirical findings more seriously. This doesn’t appear to be much of a problem for you as I know you’re already a big fan of cognitive science and whatnot. However, it is almost entirely ignored in the domain of continental philosophy. I suspect that taking seriously these findings would very significantly change the nature of the questions we as philosophers would be asking.
LS: “I think that one notable difference created in practice is that we, as philosophers, will take empirical findings more seriously. This doesn’t appear to be much of a problem for you as I know you’re already a big fan of cognitive science and whatnot. However, it is almost entirely ignored in the domain of continental philosophy.”
Kvond: I like this point very much because it goes to something important. But unfortunately the same would be the case of one merely “believed” the distinction you were making were true. It’s actual truth doesn’t seem to have much bearing. Is then your claim, it would be much better for Continentalists if they believed that there were some corresponding facts of the world that made their statements true or false? I’m not sure about this as a point of necessity. I mean, Latour seems to take the results of science quite seriously without having to presuppose a “universal knowledge”.
Mikhail,
I think your example gets at a fault-line between the two of us. You write:
Now maybe you have just chosen this example to illustrate your point, but I’ve noticed that you often speak in the phenomenological register when discussing these issues. For example, in the passage quoted above you speak of “looking”. Now, in my view, the phenomenological register is purely correlational. In the phenomenological register, we find not things-in-themselves but objects as they are for-us. As a result, I don’t think we find primary qualities in the phenomenological register.
The reason that I don’t think we find primary qualities in the primary register is because it seems fairly evident to me that we either require technological instruments to encounter these properties or mathematical formalizations. As Nick points out above, both relativity and quantum mechanics required new forms of mathematics that are quite remote from the mid-range “objects” of our phenomenological lived experience. Likewise, I have often referred to the Haldron super-collidor and others like it, which require very complex computer programs to even track the data. When I use the term “observation” I am not, generally, referring to the phenomenological act of looking (because I think the evidence is pretty strong that the phenomenological level of experience is a cognitive construction), but to all these techniques involving measuring, technology, and experiment. Since you want some criteria, I’ll provide some minimal criteria following Alexei:
First, the discovery of primary qualities requires the construction of highly controlled experimental environments that eliminate intervening factors in the phenomenon to be observed. This is why, in the detection chamber for neutrinos, scientists have built a chamber filled with incredibly pure water and highly sensitive sensors well beneath the earth surrounded by lead so as to minimize the presence of other particles, so as to capture a neutrino or dark matter.
Second, the discovery of primary qualities requires independent replicability among scientists. It is fashionable to believe, for example, that if you place headphones on your pregnant stomach playing classical music and play classical music for your newborn infant, you significantly increase that child’s IQ. This belief, known as “The Mozart Effect” was based on research done in the early 90s, and spawned the Baby Einstein phenomenon with children’s shows designed around classical music. The problem is that the observations have never been replicable, so there is significant question as to whether or not it describes a real phenomenon. The fact, however, that researchers, independent of one another, produce the same measurements of the decay of radio isotopes indicates that they are discovering something real in their measurements. Note here that while radioactive isotopes are something that we can measure through various technologies, it is not something we can observe phenomenologically. It also requires complicated mathematics pertaining to atomic physics which is at odds with much of our phenomenological experience of the world.
Finally, third, and closely related to my second point, the discovery of primary qualities requires experimentation. When I refer to experimentation I am not referring to the postmodern concept of “trying things out”. I am referring to the laboratory context where we make predictions, set up controlled environments where we diminish, as far as possible, confounds and allow the real to speak rather than allowing theory or intuition to speak. By the “real speaking” I am not talking about anything mystical, but am instead talking about the prediction/dis-confirmation relation in the laboratory setting. We painstakingly develop a theory or a hypothesis, do the extensive collective work of setting up a controlled environment, and find results that are completely at odds with our expectation. The real has spoken here and has dis-confirmed the predictions. Back to the drawing board. Note that I emphasize the term “collective”. This is because I hold to the thesis that knowledge production/discovery is not an individual affair, but a collective project that requires the work of many persons as well as the findings of history to take place. Einstein, for example, was not some isolated genius, but was a guy that noticed certain surprising features of the work of Michelson and Morley with respect to the constancy of light regardless of how fast we’re moving (speed being relative in the case of all other movements– i.e., if I’m running away from a car moving sixty miles an hour at the rate of twenty miles and hour the car approaches me at the speed of 40 miles an hour —except electro-magnetic waves or light which always moves at about 186,000 miles per hour regardless of how quickly I’m moving).
The reason I’m cagey about defining p-properties beyond the claim that they are in the things themselves is two-fold. First, were I to define p-properties I would immediately be guilty of the circularity Alexei outlines in his recent post on materialism where he argues that materialism is an idealism. That is, philosophy would be determining matter in advance, thereby rendering it ideal. Here I refer you to the work of Brassier and Laruelle as a promising avenue beyond this deadlock (the argument is far too complex to recount here). Second, I don’t think it’s the job of the philosopher to tell us what the p-qualities are. These are discoveries made by scientists. This is one of the reasons– in a statement that recently made your jaw drop –that I said it’s not the job of philosophers to tell us what time and space are. Yes, I agree that philosophers can say all sorts of interesting things about the phenomenology of time and space. But as for space as real, we are ill-equipped to say. As a meta-theoretical discourse, both you and Alexei have already agreed that philosophy doesn’t impinge on science, so this shouldn’t be objectionable to you… Unless you believe we know before we know.
You write:
I think the foregoing comments outline why I do not believe my claims is based on an “intuition”. But in addition to this, my argument that we know some primary qualities has not been based on an “intuition” or “hunch” either. It has been based on the rather non-controversial thesis that for everyone, save the crazy creationists, evolution and the Big Bang have told us something true about our universe. I have endorsed Meillassoux’s argument that these claims are only intelligible within the framework of a realist epistemology and metaphysics. Above I have outlined some very broad criteria as to how these p-properties are distinguished from s-properties (that is, I’ve provided the rudiments of a “meta-theory” of knowledge of these properties). And you’ll note that in outlining this meta-theory the possibility of falsification is everywhere present.
Kvond,
Sure, I’m willing to concede that a “belief” would do the same work. The problem, I think, is that there are a number of things internal to most continentalist philosophical assumptions that prevent this belief from getting off the ground. You seem concerned with the problem of dogmatism or the realist believing he has attained the absolute. In my case, I’m willing to go with the much more modest assumption of inductive probability, i.e., that given such and such evidence this is the most likely explanation. For example, given all of the evidence of neurology, the most likely explanation is that the mind is the brain. I’m more than happy to concede that– especially given the infancy of the science –we’re probably wrong in a number of the details and that we have a lot left to explain. The problem in the continental circles is that these discoveries aren’t even granted admittance to the door at SPEP or the Society for Phenomenology and Existential Philosophy conferences.
As Ben recounted– and I’ve often experienced on my own and even argued on my own –when sitting between the advocate of Jean-Luc Nancy and a Kantian, he’s told by the one that science is equivalent to faith and by the other that science is purely speculative. That’s crazy! What set of philosophical assumptions are at work here? This entire logic haunted a thread that went to some thirty or forty comments between Mikhail, Alexei, and I. Alexei kept arguing that Kant grounds science and therefore science can’t discover anything that contradicts Kant. As a result, Alexei authorizes himself to ignore anything that might be discovered by science or mathematics a priori because he begins from the tautological premise that the sciences can never contradict their transcendental grounding. As a result, he “knows before he knows” in the sense that because he already has the so-called “conditions” in hand, he can ignore all this dirty empirical stuff. I believe you and I saw the “cash-value” of this premise in our discussion here about normativity, where we were told that neurology or the naturalistic approach could never, under any conditions, call into question the Kantian ethical philosophy despite the findings of Damasio regarding affectivity and moral reasoning. Transcendental argumentation, despite its intimidating arsenal of concepts and arguments, ultimately becomes a defensive fortification that functions to protect, as Hume would say, superstition from empirical observation by dressing itself up in fine robes.
Maybe you could say a bit more regarding Latour and his respect for science. When Latour says that claiming that explaining the Bubonic Plague as caused by bacteria is an anachronism, it seems to me that there’s a problem with his ontological relativity. What, exactly, do you have in mind by your term “universal knowledge” and how am I positing universal knowledge?
LS: “You seem concerned with the problem of dogmatism or the realist believing he has attained the absolute. In my case, I’m willing to go with the much more modest assumption of inductive probability, i.e., that given such and such evidence this is the most likely explanation.”
Kvond: I like this very much, and if this is all that you claim then we would be in complete agreement. It only remains that I have no idea how this is served or reached by some kind of Realism. One simply doesn’t need a Realist claim to take this position.
LS: “Maybe you could say a bit more regarding Latour and his respect for science. When Latour says that claiming that explaining the Bubonic Plague as caused by bacteria is an anachronism, it seems to me that there’s a problem with his ontological relativity.”
Kvond: All I said was that he took science seriously. I think he is very serious in how he engages science. I too have a problem with how he thinks about relativities, seeing truths as politically made so to speak. But he certainly takes the products of science seriously. Perhaps you have in mind some kind of weightiness, a certain kind of “respect” that should be paid.
LS: “What, exactly, do you have in mind by your term “universal knowledge” and how am I positing universal knowledge?”
Kvond: Hmmm. I took the phrase from what you wrote above when qualifying certain thinkers as the worst because they did not attempt to account for something like “universal knowledge”
LS wrote: “Now were I to grade anti-realists from best to worst, I would place figures like Kant and Husserl towards the top of the list as best and figures like Lyotard, Wittgenstein, and Baudrillard among the worst. At least in the case of Kant and Husserl we get an attempt to account for something like universal knowledge, even if it is only ever restricted to phenomena. When we get far down the correlationist spectrum to folks like Lyotard, late Wittgenstein, and Baudrillard, all we get are competing “world-versions” that are unable to claim any superiority over one another.”
Levi writes,
“Now, in my view, the phenomenological register is purely correlational. In the phenomenological register, we find not things-in-themselves but objects as they are for-us. As a result, I don’t think we find primary qualities in the phenomenological register.”
This reminded me of another point that problematizes the approach being taken here by Mikhail and Alexei – namely that the demand for the criteria that would divide p-properties and s-properties seems premised on ascribing properties to objects. It’s not clear, however, that our standard intuition of objects holds up in any meaningful sense once we try and exit from the phenomenological. So for example, there’s an entire movement in analytic philosophy of science called structural realism. One of it’s main points is that what science picks out as real is mathematical structure (and not individuals that can be predicated of!). For instance, while Newton’s specific theories are discredited (beyond certain limits), the mathematical structure ingrained in his theories has managed to be retained through special and general relativity. So while general relativity is almost certainly wrong in some key aspect, structural realists argue that historical experience teaches us that the mathematical structure will remain – and this because it is knowledge of the real.
I’m not sure where I stand in regards to this structural realism (I’ve only begun looking into it), but it seems important to highlight that speaking of p-properties and s-properties might be misleading (and already conceding too much to the phenomenology of experience).
Ack! I meant Maxwell’s work on electro-magnetism, not Michelson- Morley. Such are the dangers of responding to comments late on a Saturday night.
Nick, could you cite any of the structural realists. I confess I’ve grown a bit suspicious of mathematical formalist realists in light of Meillassoux’s arguments and after reading Catren’s article. It is not that I disagree with the thesis that mathematical formalization is crucial, it’s that I think it invites mathematical speculation per se that is then transposed on to the natural world. I think it’s important to preserve the distinction between mathematical logico-deductive necessity and natural necessity, where the latter refers to laws that are contingent (non-deducible) but nonetheless operative in our universe, i.e., if x had not happened, y would not have happened. Maths, as I understand them, posit no such limitation or restriction, allowing whatever follows from a formalism to, well, follow from that formalism. I think this is how a lot of Badiou’s arguments work. Essentially he claims that because there is no ordering principle in a set, being has no ordering principle. But this strikes me as a metaphysical leap. The fact that the elements of a set can be ordered in any old way doesn’t entail that the natural world, linguistic world, or social world can be ordered in any old way (granting that between these three levels there are varying degrees of ordering). This strikes me as exactly the way that Meillassoux’s arguments about contingency function. I guess I’m going back to an old theme here that I’ve been obsessed with since the earliest days of Larval Subjects: Is mathematical formalism sufficient to capture being-existence in its claws?
LS: “I guess I’m going back to an old theme here that I’ve been obsessed with since the earliest days of Larval Subjects: Is mathematical formalism sufficient to capture being-existence in its claws?”
Kvond: What comes to mind is Giambattista Vico’s view that mathematics is the closest man comes to being God in that man makes maths out of nothing, and their truths are truths that are completely accessible simply because we have made them (much as history is accessible insofar as we have made it). This follows his verum-factum principle: the truth and the made are convertable (verum et factum convertuntur). He also claims that for this reason we cannot ever “know” creation, because we did not make it. I think he makes a very good point or distinction.
What I would want to know, in my usual style, is what is the difference if it does catch it in its claws, or doesn’t? This ultimately is a question of authority.
Levi, I’m reading through James Ladyman and Don Ross’s Every Thing Must Go right now. I agree with your reservations about mathematical structure, and while that’s apparently the typical form of structural realism, Ladyman and Ross argue for a somewhat different variant. They take ‘real patterns’ to be what exists, with these patterns being defined in terms of informational – rather then simply mathematical – relations. (I don’t claim to understand their notion of information very well though. It involves a fair amount of computational theory, so I need to do some background research first.) They even explicitly argue that mathematical and logical possibility is far too expansive, and that physical possibility should be the determinant of what exists. You might also like it since they incorporate what they call ‘rainforest realism’ which is akin to the principle of irreduction. Basically real patterns don’t exist solely at the level studied by physics, but also at higher-up levels (e.g. at the level of natural selection, or traffic flows). These patterns may eventually be reducible to a different pattern, but that’s always an empirical matter. Really interesting book though, even if I don’t agree with their reduction of philosophy to being a handmaiden of science!
Levi (and Nick, welcome to the philosophical soiree), I think I like the turn of the conversation, we are moving definitely somewhere here, my example was indeed indicative of my approach to the problem and your response was intriguing.
Let me get my thoughts together and recover from an awesome BU vs. Miami (Ohio) NCAA championship game I have had a privilege to witness this evening (alas, only via a television) – what a goal, friends!!!
“It’s not clear, however, that our standard intuition of objects holds up in any meaningful sense once we try and exit from the phenomenological.”
That’s my sense of what’s been missing from what I’ve followed of this back and forth. I don’t know about mathematical structure, but once you move to a mathematical understanding of nature in terms of laws and events or terms and relations, you’re no longer talking about things or substances. In Kant, I believe, the thing in itself could be an apple or the entire universe, I don’t think Kant ever comes clean on that, but the Kantians can correct me, I assume.
Interesting discussion and I’m intervening rather late, so I’ll weigh in broadly. Doesn’t the realist simply claim this:
The P/S distinction is certainly justified I think if we simply appeal to our natural abilities to sort out the primary from the secondary, and we do this instinctively, I think. In fact, isn’t this distinction –as someone pointed out above–irreducibly relative to us? That is, to our biological/bodily and mental life and with regards to the particular state of affairs we find ourselves in. Mikhail’s question regarding how do we know that, say, shape is one of the primary qualities of a pretzel and the salty taste is one of its secondary qualities, begs some important metaphysical and logical questions.
One thing we could do is test it, right? When Mikhail and I get together and eat some pretzels we are generally presented with a sensory experience that we know is the very experience of “tasting.” It is built in the nature and quality of this particular experience to inform us that what we are experiencing is an appearance. Shape is a different case. Clearly, we have sensory experiences of primary qualities, And there are certainly criteria independent from Mikhail and I’s pretzel eating experience/party that allows us to decide whether a primary quality like shape is, I don’t know, “tangible” or if it merely seems tangible. Yet, for SQ the distinction between real and apparent saltiness have to be derived from our experience of eating the pretzels. This is not the interesting part. What is rather interesting is how we make this distinction between the ways in which the real and the apparent “giveness” of qualities is to be drawn is obviously, crucial. Certainly, one tack we may take is this: we rest our judgments pertaining to whether an object possesses a primary quality may be justified by means which are not based on the experience of an appearance, yet, PQ have an appearance. So, our experience of PQ may not be part of the justification for our judging that an object possesses them, whereas our experience of SQ must be part of the rationalization for our judging that an object possesses them.
While the experiential and non-conceptual content of our sense data/experience has an epistemic value, it’s quite conceivable that we could not grasp either the content or this epistemic value. So SQ may well be true without evidence, e.g. without any recourse to experience and non-conceptual content. So here’s a question: Why not abolish primary qualities altogether?
A lot has happened since the last time I commented, so I’m bound to let a few threads drop here.
I did note, however, a potentially unhelpful equivocation in some of the things I said earlier, regarding knowledge. I assimilated the properties of knowledge to the properties of a scientific theory too quickly. This said, I don’t think it really changes much of what I said, but it’s worth pointing out.
anyway, I want to address a few of Levi’s responses to me and others. First, this one:
Well, in the first instance, and based on what you’ve explained to us in the past, I don’t think Bhaskar has actually formulated a fallacy — especially since it really amounts to saying in some cases it’s false to always reduce ontology to epistemology; hence in some cases it is perfectly legitimate to do so; If you want to claim anything stronger, you actually do need to provide the conditions for false reductions, which I haven’t seen. At the moment, then, the so-called epistemological fallacy is vacuous.
And I don’t think there’s anything specifically wrong with saying that ontological commitments are coherent only in light of epistemological ones. I mean, I can be ontologically committed to the existence of the 108 Gods of Voodoo, but this commitment is totally incoherent on epistemological grounds, because supernatural forces have been eliminated from playing an evidential role; independent of the grounds of knowledge, my ontological commitment is nothing more than faith. And the more I think about it, the more I’m inclined to say that there seems to be a fundamental slippage in your notion of realism, Levi. So normally we tend to think of realism in terms of these basic theses:
Although a little reductive, realism simply is tantamount to (1) a relationship between concepts (predicates) and properties (anti-realism is simply an attempt to complicate — add nuance to — this relationship). Notice, however, that this one metaphysical claim about the independent existence of properties and states of affairs functions to justify or lend support to a particular conception of truth and justification and therefore serves as part of a metatheory (justification of the justification, theory of the theory). One of the reasons the positivists thought that any sort of metaphysical claim was totally foreign to science was precisely because they realized that contexts of discovery aren’t coextensive with contexts of justification.
Moreover, a metaphysical claim is only probative when one has a particular — logically antecedent — theory of truth and justification that one wants, in turn, to justify. IN effect, all (anti-)realisms are little more than a specific set of beliefs concerning truth and Justification. What makes theory of justification 1 more coherent than theory of justification 2? — the fact that ‘truth’ is a correspondence between state of affairs and proposition. Realisms and anti-realisms stem from epistemological considerations (what is knowledge justification, truth etc.). But to the extent that this statement is true, we’ve already conceded the basic point of contention between us here, namely that our ontological commitments in general — and the very distinction between primary qualities and secondary ones — are only coherent within a given epistemological framework.
Now, about this:
I don’t think that any of this follows at all. What I would say it this: were we in possession of a true transcendental theory, nothing could contradict it — that’s just the upshot of transcendental philosophy, and it’s no different from any other kind of true theory; nothing in experience contradicts the theory of arithmetic, even though one drop of water added to another drop of water doesn’t equal two drops of water (one flame added to another produces one flame, and the list continues…); I mean really if you’re in possession of a true theory, then contradiction is impossible (unless you think that some contradictions are true). But from this it doesn’t follow that one knows before one knows — i.e. that empirical research isn’t important. At best, one knows something about what determines ‘knowledge,’ but that doesn’t impinge on actual discovery or increases in knowledge via empirical and pure research. Nor does it means that a transcendental philosophy can’t be shown to be false by empirical science. It’s just not as easy as you think. So really, your whole line of criticism here kinda misses the point.
This is not to say, however, that Kant gets it right, or that some aspect of contemporary science doesn’t contradict it. And I’ve never claimed that. I have claimed that none of the examples you’ve brought up contradict what Kant has argued. But that’s a different story, and has a fair amount to do with what counter-examples can actually demonstrate. Simply put, it’s cheap and unproductive to say things like, ‘see Relativity shows Kant’s wrong about space.’ That may well be true, but unless you do the hard analytical and argumentative work, you might as well be saying relativity proves Kant right, and also shows that the moon is made of cheese. That is, so far, it’s just an assertion.
All this said: I still see no reason why I should be motivated by realism, I still don’t see what swath of hard won knowledge anti-realism deems illegitimate, and I don’t understand why you think ontological commitments can be independent of epistemological ones, without thereby loses any rational ground for commitment.
Shahar,
I’m not entirely certain that I understand your argument in your most recent post. At any rate, three points in response:
First, as always I think it’s tremendously important to distinguish the epistemological and the ontological. I think the question of how we know is a tremendously important question for both the realist and the anti-realist.
Second, the realist (in my version, Graham’s different I think) is not making the claim that “p is true regardless of whether or not we have evidence.” Truth is a relation that pertains to knowledge. The realist will claim that “entity x has property p regardless of whether or not humans exist” (an ontological claim), but this claim is an entirely different claim than the claim that we are warranted in asserting the existence of a property without any evidence. In order for us to be warranted in asserting the existence of any property, we have to be capable of providing evidence, otherwise we are simply making pie and the sky speculations about the world about us.
A while back I wrote a couple posts on the Roy Bhaskar’s distinction between the intransitive and the transitive: (http://larvalsubjects.wordpress.com/2009/02/02/roy-bhaskar-transcendental-realism-and-the-transitive-and-the-intransitive/ and http://larvalsubjects.wordpress.com/2009/02/02/roy-bhaskar-transcendental-realism-and-the-transitive-and-the-intransitive/). Perhaps the language of this post was difficult to follow, but the point was to draw a distinction between the historical dimension of inquiry and the ontological dimension of beings. The transitive refers to shifting theories and bodies of evidence, to the role politics and culture plays in scientific theorizing, etc. The intransitive, by contrast, refers to those properties that belong to objects independent of our theories. The aim is to discover the intransitive, the claim is that the intransitive possesses whatever properties it possesses independent of our knowledge of these properties, but we cannot assert specific properties without the providing of evidence.
Here Alexei’s observation of String Theory is to the point. It is indeed a mathematically elegant theory that unifies QM and Relativity, but as of yet there is no evidence that it is true. Now I disagree with the scientist that Alexei cites and his assertion that working out the details of this theory are a waste of time. It took a long time for both Newton and Einstein to be confirmed. Working out the details of an uncomfirmed theory plays a role in devising experiments that will provide that evidence that will allow us to determine the truth of the theory. However, until that evidence is available we would be falling into dogmatism in claiming that “atoms posses property p despite the fact that we have no evidence for it.” We just don’t know one way or another at this point.
Finally, third, and most importantly: At the level of epistemology, the debate between the realist and the anti-realist revolves around the Kantian distinction between phenomena and the in-itself. Were Kant simply claiming that something must be given to us to know it and that the object, event, or thing that gives itself has this property regardless of whether or not we observe it, there would be no disagreement between the Kantian or correlationist and the realist. The debate between the correlationist and the realist arises when the correlationist adds “we have no idea whether the thing-in-itself has this property and therefore can only say that this property is for-us, not in-itself.” I’ve repeated this point so many times, yet it still seems to get lost in the discussion. The debate isn’t about whether or not humans must relate to the entities they investigate to know them, but about whether or not the relation constitutes those entities. The realist says decidedly not! The anti-realist says that objects (not things) are what they are by virtue of how the correlation constitutes them such that we can never know whether or not things have these properties. That’s where the dividing line is.
Alexei,
Allow me to give a few examples of conflating the epistemic and the ontological to indicate why I think such a move is a fallacy. Suppose we take Hume’s critique of causation. When Hume claims that our knowledge of causation is grounded in relations of sense-impressions he is making a claim about how we come to know causation. Causation, claims Hume, is a constant conjunction of sense-impressions. The epistemic question then becomes how we are able to ground our knowledge of the necessity asserted by causal claims. Fine, this is great as an epistemic question. However, the ontological question of causation is not a question of how we know causation, but a question of what causation is. In fact, Hume’s epistemological critique of causation is only intelligible on the grounds of an ontological account of causation: that it involves necessary relations between entities, that it “powers” within entities, etc.
Things get hopelessly muddled when we don’t keep these two levels distinct. Thus, for example, a match ontologically has the power to light a fire regardless of whether or not we observe this sequence of events. Moreover, it is not clear that actualized cause and effect relations involve the production of sequential sense-impressions as Hume suggests. The food I eat plays a role in the production of my cells, but I don’t observe the bundle of sense-impressions embodied in my delicious pasta and the causal interactions taking place throughout my digestive system as this pasta is digested. Here we have the perfectly legitimate question of how we know this causal interaction is taking place, and the ontological claim about this causal interaction itself that is independent of our knowledge of it.
We see a similar slippage between the epistemological question of causation and the ontological status of causation in discussions of quantum mechanics. Heisenberg’s Uncertainty Principle tells us that we can only ever determine the position or momentum of a particle at any given point in time, but never both at once. Heisenberg’s claim is an epistemic claim, not an ontological claim. As a result of the failure to carefully distinguish between what is being claimed epistemically and ontologically, we get all sorts of people claiming that the universe is indeterminate, because they transpose a limit to our knowledge into a property of things. It may indeed be true that there is an intrinsic limit to our knowledge when dealing with quantum phenomena, but nothing in Heisenberg’s formulation of this limit warrants the leap to the ontological claim that quanta are indeterminate. They might be, they might not be.
Finally, third, there is the case of Kant. In my view, Kant fundamentally misunderstands Hume’s problem. I can agree with Kant that we are at a loss to understand how we can ever have the concept of necessary relation based on sense-impressions and principles of association. No matter how often two events are conjoined with one another they will never produce the idea of a necessary relation. Kant “solves” the problem by arguing that our idea of necessity comes from mind or world, but if so many people have been left scratching their head over this “solution” then this is because it doesn’t really get at the root issue of how we distinguish between those relations among things that are truly necessary and those that aren’t. When I was growing up my parents had a handsome set of leather bound encyclopedias from 1910. Upon looking up asthma one day the encyclopedia authoritatively informed us that asthma is caused by constipation. Now, under a Kantian construal, the mind had applied the category of cause and effect, asserting a necessary relation between these two events. But here Kant is no better off than Hume. Just as Hume argued that we have no way of distinguishing between the merely correlational and the causal, for Kant, despite the assertion of necessity, we run into the same problem. This problem, in part, arises from bracketing the ontological dimension of the issue and addressing it only at the level of the epistemic or cognitive.
Levi, all this stuff about Hume and Heisenberg is well and good, but it doesn’t seem to address anything I’ve said. For Hume explicitly begins with an epistemology (theory of impressions, theory of ideas), and then moves to an ontological question.
so again, I fail to see how one can even pose the question, ‘What is x’ without some antecedent framework, which makes ‘x’ a meaningful entity. this is really a basic point (a kind of hermeneutical issue writ large), so I’m not sure why you’re resisting it so much. It doesn’t mean we can’t be totally wrong in our initial conception of what an entity is or what it means to be an entity of type ‘x’. It just means that it makes absolutely no sense to talk about entities absent a meaningful framework.
I really don’t see why this claim chafes you so much. But let me reiterate my questions, since I still don’t have an answer to them: why should I be motivated by realism, what swath of hard won knowledge anti-realism deems illegitimate, and why you think ontological commitments can be independent of epistemological ones, without thereby losing their rational ground for commitment?
LS: “The debate isn’t about whether or not humans must relate to the entities they investigate to know them, but about whether or not the relation constitutes those entities. The realist says decidedly not! The anti-realist says that objects (not things) are what they are by virtue of how the correlation constitutes them such that we can never know whether or not things have these properties. That’s where the dividing line is.”
Kvond: What I am curious about is, what is the difference for you between saying “knowing whether…things have these properties” and “it is quite useful to act as if things have these properties”. Your earlier assertion that you merely want to claim that your knowledge of things is highly probable knowledge seems to point toward the latter, but indeed you seem to want something extra, a little bit of a tip of the cap “reality” as an ultimate “buck stops here” point, things REALLY have these properties. I mean, really, really, really have these properties. Your position seems to be a point of emphasis, rather than a full claim.
Nick,
I think that QM basically follows the critique of Kant launched by Fichte (and then Hegel) – to put it simply: to say that things-in-themselves somehow cause our things-as-they-appear is to apply the category of causation outside of the limits of allowed space/time experience, therefore [insert your favorite German idealist philosophy]
I might be mistaken but I don’t recall QM doing anything different.
Pingback: Experience: Now With New (Scientific) Sause. « Perverse Egalitarianism
Alexei,
You don’t seem to be reading me very carefully because I have agreed with you on this point repeatedly! I’m beginning to feel that this discussion is a bit like a discussion where someone asks “why do you want x?”, the person responds “I don’t want x!”, and then the first speaker says “yes, but why do you want x?” What you are referring to as the hermeneutic dimension is what I have called the “transitive”. Yes, yes, yes, I agree! Our observations of the world and experiments occur within a hermeneutic horizon of meaning.
I have been exceedingly clear– multiple times ad nauseum –as to what I reject in the correlationist approach: the thesis that we only ever know phenomena and that we can never know whether or not things-in-themselves possess these properties. Full stop.
I have written three posts on Meillassoux (four if you include my post “Time, Turtles, Kant, and Correlationism”) carefully explaining what swaths of science I believe the correlationist thesis as described above with respect to phenomena cuts out. Please read them. I believe it undermines any knowledge claims about the world prior to the emergence of life.
I think you guys keep missing or eliding this central Kantian thesis about phenomenality and thus miss what this debate is about.
For the record, I think pretzels are an abomination and an embarrassment when it comes to foods – I am very openly anti-pretzel, no doubt.
I think that’s my whole point in these discussions, I don’t think the language of “primary qualities” is something I really care for, I am willing to hear out the arguments yet I am not really persuaded that Meillassoux’s call to bring the distinction back is that attractive…
If we agree Levi, then we also agree that ontological commitments are coherent only light of epistemological commitments.
Do we still agree? If so, then it also follows that whatever we take to be a p-quality depends upon the epistemological framework we’re using, since the distinction between p- and s-qualities is internal to this framework (although this framework depends upon previous scientific research). Hence it’s not obvious that p-qualities have anything to do with things in-themselves. And this implies that you’re just as anti-realist as I am.
Now, I’m sure we don’t agree on this last point. And I’m trying to figure out why. It strikes me that the moment you accept the hermeneutical point, you can’t deny the anti-realist/Kantian interpretation of things in-themselves (i.e. things in themselves are unknowable and posited by a theory in order to delimit possible knowledge claims [true in the theory T] from non-knowledge). Put differently, if being that can be understood is language, and the thing in itself isn’t language, then it can’t be understood.
I think you are overestimating the power of that argument – it’s not as solid as you think it is in itself, but even if it was, it’s not as powerful as you think it is – you always bring it up as though it is a kind of powerful tool in destroying any other correlationist argument which is not so. As I said before, I am not at all persuaded that QM’s argument is even an argument – you did write three posts on QM, but you basically restated his argument which, for those who actually read the book, is not really saying much. That is, I appreciate your summaries of QM but you are not really adding anything to the essence of his argument. I still think that his response to the imagined rejoinder about how ancenstral is the same as non-seen is pretty weak. We did touch on that – all that business about “time before time” – “givenness” etc etc – what reads like a tight argument to you is actually not so tight to me…
Alexei,
I guess that your argument here strikes me as based on a sort of paralogism. I don’t see how the thesis that our investigation of the world takes place within a hermeneutic horizon– say a Kuhnian paradigm –leads to the conclusion that this knowledge produces an object rather than yielding access to a thing.
Mikhail,
Does Kant, or does Kant not directly claim that we can have knowledge of whether or not things-in-themselves are spatially and temporally organized as our experience is (hint: he does). Given this, how can the claim that human consciousness, the source of these forms of space and time according to Kant, emerges from a time prior to human consciousness, possibly be intelligible for the Kantian? If you claim that this is intelligible you’re claiming that the forms of time and space are operative and directly applicable to things-in-themselves prior to the emergence of mind, which is absurd within the constraints of Kant’s own epistemological theory.
Mikhail,
“I think that QM basically follows the critique of Kant launched by Fichte (and then Hegel) – to put it simply: to say that things-in-themselves somehow cause our things-as-they-appear is to apply the category of causation outside of the limits of allowed space/time experience, therefore [insert your favorite German idealist philosophy]
I might be mistaken but I don’t recall QM doing anything different.”
I’m glad you replied to this because it’s something I’ve been wondering, following these discussions. Specifically, I don’t think this is QM’s argument (or rather it’s only one side of it). So what QM does is set up the argument as a 3-way debate between a weak correlationist (Kant), a strong correlationist (Hegel), and himself. In order for the weak correlationist to fend off the absolutization of the correlation, she must invoke the facticity of the correlation. (One cannot uncover the logic of the transcendental conditions, but can only describe them.) But in doing so, she inadvertently reveals that she already has knowledge of the absolute (namely the necessary contingency of everything, including the current transcendental conditions).
So when I say that the weak correlationist is in an unstable situation, I mean that either she must accept Fichte and Hegel’s critique and admit that the noumenal is an unwarranted metaphysical residue. Or to avoid that, she must invoke the facticity of the correlation, but in so doing, take up QM’s position.
Maybe I’m reading this wrong though, but if I’m right, I haven’t seen any weak correlationists try to refute it yet.
Nick,
I think that if this conversation went into the discussion of post-Kantian criticisms and possible Kantian defenses against such critiques, I would certainly have found myself in a more comfortable position as it would deal with more historical issues. However, the main reason I’ve enjoyed the recent exchanges was because I found myself out of the usual loop and I think I am thinking about familiar issues now from a rather new perspective. If you are, however, interested in that particular debate, I would gladly partake in it – maybe you can post something about this and I will respond and we can get the ball rolling.
My first reaction would be say that Hegel’s critique of Kant is more eloquent than Fichte’s (it seems to me), especially the famous “to know the limit is to cross the limit” argument – that my phrase, Hegel actually uses slightly different expressions [its in Science of Logic, in Determinate Being chapter, there’s a “Remark: The Ought,” I think in the B. Finitude section – too lazy to look it up now].
However, to set up the threesome the way you think QM does it is tricky – it’s to assume that you can do a kind of reductio ad absurdum move that Levi likes so much and thinks its powerful enough to crush all correlationism forever – that is, if you show that weak correlationism does not hold, it must be true that either strong correlationism is correct and then by showing that it cannot account for statements X you destroy it as well, then the only option left is yours (QM’s in this case) – I don’t know if I would agree with that. Proving that idealism is wrong does not make one a realist, does it? [See Kant’s refutation of idealism, yet we can hardly think of him as full-blooded realist according to Levi’s definition]
Definitely agreed – I don’t want to sidetrack this great discussion. It’s something I haven’t seen referred to by anyone though, so I was really curious since I think QM tries very hard to grapple directly with all the Kantian-type questions arising in these discussions. I’ll try to draw up a separate post on it, although my mind has been focused on other topics lately, so who knows when I’ll ever get around to it. Hopefully soon!
Nick, you don’t have to if you don’t have time, but all I am saying is that I would love to discuss that issue separately, I just don’t want to start a new thread here…
Again, no offense, but your excellent refutation of Kant is based primarily on a peculiar misreading of what Kant has to say about space/time – they do not come from consciousness, they are forms of intuition which is not the same thing, of course.
Things are intelligible for Kant if thinking them does not involve contradiction – if you claim that human consciousness creates space/time, then of course thinking about a “time before time” would be contradictory, but of course is that really what Kant does? (hint: no, it is not)
I still don’t understand what primary qualities are.. If one of you can please help me out on that and with an example because I’m not quite getting it.. Thanks
Mary, I always thought that the distinction between primary and secondary qualities came from Locke (in this formulation), but it seems that Meillassoux want to go back to Descartes – for Locke’s discussion of see An Essay concerning human understanding, Book II, Chapter VIII – he has some good examples there as well.
I haven’t thought about Kuhn in a few years, so maybe I’m wrong, but I’m confident that everything I’ve said is quite consistent with his views (save maybe for my appeal to contexts of discovery and contexts of justification but that’s a separate matter), and doesn’t produce any sort of paralogism. Furthermore, his realism strikes me as critical rather than speculative.
Take a look here. Here’s what Alexander Bird has to say about the matter:
And,
So anyway, I’m not alone in thinking about Kuhn’s work in this way.
Finally, I think it’s a mistake to think of what I’ve said as constituting an object; all I said is that objects are determined in the way I’ve described. There’s a world of difference between the two.
Whoops. Sorry, Levi. I just read my comment above. Not that it matters now, but instead of “logical and metaphysical” it probably should have said something like modal and occurrent/dispositional instead, based upon those who cut up P and S Q’s further. Mikhail’s example of a list of qualities raised this kind of question for me, but I was too distracted to make any kind of coherent comment. So much for watching Sunday am tv and posting comments at the same time…Ack.
No need for the bulleted lecture, however. Just say “Shahar you whelp, what the hell are you saying.”
Oh, and Mikhail, very sorry to hear you’re back on an anti-pretzel platform.
Shahar, my anti-pretzel platform is as strong as my previous anti-popcorn platform – I wonder if I will ever qualify for an American citizenship with such obvious anti-American views…
Nick,
You said:
“So what QM does is set up the argument as a 3-way debate between a weak correlationist (Kant), a strong correlationist (Hegel), and himself. In order for the weak correlationist to fend off the absolutization of the correlation, she must invoke the facticity of the correlation. (One cannot uncover the logic of the transcendental conditions, but can only describe them.) But in doing so, she inadvertently reveals that she already has knowledge of the absolute (namely the necessary contingency of everything, including the current transcendental conditions).”
Actually, in QM’s vocabulary, “strong correlationism” is not the same thing as “absolutization of the correlate,” which for QM describes Hegel (as well as Schelling, Deleuze, Nietzsche [cf. AF 37]). Strong correlationism (i.e. Heidegger, Wittgenstein) is defined by the rejection of this absolutization by affirming the facticity of the correlate.
QM’s own argument against strong correlationism parallels Hegel’s argument against Kant, by absolutizing facticity as Hegel had absolutized the correlate (cf. AF 51-2).
I have to say, I really liked Levi’s up front answer to the question as originally posed:
Levi: “I’m afraid that I just don’t have an answer to this question at this point. That is, this is a philosophical question and problem to be worked out and solved, not an answer that I have in hand.”
In some ways, this is really what the process of philosophizing is all about. Its turning all the variables, and then seeing if you missed one, or mislabeled one (or all).
Thanks Malevich – I noticed that as well when I went back and looked at the book. It’s a bit more nuanced than the trialogue that I’d set up (e.g. there’s skeptics and dogmatists introduced too). But I think the general idea holds up – that in the Mexican stand-off between all these positions, it’s QM’s position that’s the only one left standing at the end (at least, that’s the argument – who knows if it holds up). I’ll have to flesh it out more in a future post. (And Mikhail, I’m interested in the question itself, so I’m happy to try and articulate my own thoughts on it. I’m not sure about this anti-popcorn platform though; that just seems borderline inhuman…)
Nick, I mainly object to the smell of pop-corn, at least, that’s my official version – unofficial one goes back to my first introduction to this exotic American food when I first observed the magic of conversion of corn into pop-corn – big deep trauma…
I think QM’s position wins in a stand-off precisely because if it didn’t, it would have been strange that he even wrote the book to begin with…
I’d like for this angle to receive more attention and I would join right in (hopefully my technical difficulties would subside by then)
This can be a question to anyone, but especially those on the side of QM: I wonder why he doesn’t just go for the kill in terms of Kant’s position on space which is presents us with a bigger problem than that of time:
If according to QM/Levi, we cannot explain “ancenstral statements” because the mind produces time and there was no time back then before humans, then can’t we say the same thing about Kant’s idea of space – was there no space back then as well? if there was no space, because space is produced by the mind, wouldn’t we have even more problems explaining, say, distant galaxies and all of that stuff?
Once I am at it – another question: Wouldn’t Kant necessarily deny that we can know that something like a big bang ever took place since then we would know that the universe has a beginning and therefore is finite? According to the first antinomy we cannot know whether the world is finite or infinite – doesn’t this basically answer the question for Kant?
Actually, Lou, these are great questions – if time/space are indeed products of consciousness, as Levi seems to suggest Kant wants to claims, then the problem of space would indeed be something akin to what you are talking about – if space is a result of mind’s imposition (again, the kind of language that Levi uses and the kind of language I usually protest against), then there was no space before humans, right?
I actually address this in my third post on Meillassoux “Meillassoux III”. Although I think Meillassoux ultimately would argue that space is real, he doesn’t seem to think of space as having the same bite against the correlationist argument. The reason for this is that, under a classical construal, space is the domain of simultaneity with respect to time. Consequently, since distant galaxies are simultaneous with us the counter-factual argument works because consciousness has emerged at this time: “if we were to visit these galaxies they would have these spatial characteristics”. The temporal argument is different. Because they are premised on a time before life and the constituting activity of consciousness, they have the appearance of being dogmatic statements as they require us to speak directly of things-in-themselves.
So there’s “time before time” but not “space before space”? What is the “classical construal” that you are referring to? Aristotle? It seems that you are willing to accept QM’s strange “time before time” – that is what you called “real time” – yet when it comes to space, it seems strange to say “space before space”? Is there “real space” as well?
In your QM III post you say:
What I am talking about it not the argument from un-witnessed, it’s an argument from “space before space” or “space before givenness” – if, according to Kant, for example, there was no time or space before the humans came around, he must be a really strange philosopher not unlike Berkeley?
Pingback: Who’s Dogmatic? « Larval Subjects .