Ben Woodard
Nick Srnicek
Taylor AdkinsIn the midst of all the debates about realism lately, one thing the critics seem to often misunderstand is the relation between science and realism. Most significant is the suggestion of dogmatism with regards to science – as though realists blindly took current scientific theories to be comprehensive and definitive. As though realists weren’t aware, along with everyone else, that science’s history is filled with theories being posed and discarded. In addition, as anyone with a passing familiarity with science is aware, there are absolutely massive gaps in our knowledge of the world. As of right now there are two competing theories about the fundamental nature of physical reality – quantum mechanics and general relativity – yet these theories are incompatible. Moreover, it is estimated that 90% of the universe is comprised of a mysterious thing called dark matter – yet we know virtually nothing about it (nor its cousin, dark energy). These gaping holes are situated in the most scientific discipline we currently have – fundamental physics – yet they clearly point to the fact that current scientific theories are in desperate need of massive revision. (And for more examples of major holes in scientific knowledge, see this article: ‘13 Things That Do Not Make Sense’)
In the literature on philosophy of science, these problems are all well recognized. Referred to as the ‘pessimistic meta-induction’, the main argument proposed by anti-realists is that history shows that every scientific theory previously taken to be a true representation of reality has ultimately been tossed aside. Using history as our basis, the argument suggests that we should be very pessimistic about the possibility of current theories maintaining their status as truth.
As a response, scientific realists point to two arguments. One, the success of science in explaining and, importantly, predicting the outcome of experiments would have to be considered to be a miracle if science didn’t grasp onto some aspect of reality. It would be an astounding coincidence if our theories had no hold on a mind-independent reality, yet still gave us such accurate predictions. But secondly, taking into account the pessimistic meta-induction, scientific realists have agreed that there is no reason to believe that the entities postulated by our current theories are definitive and exist in some fundamental way. This sort of realism about the entities of scientific theories seems to clearly be discredited. So what scientific realists point to instead is the structure of the mathematical theories put forth by science. Whereas it can easily be shown that the entities postulated by theories in the past have been invalidated (e.g. caloric substance being replaced by the kinetic theory of heat), it can also be shown (and is an active area of research) that the mathematical relations are maintained between theories. A simple example will hopefully suffice to show what this means.
So in classical mechanics, force is represented mathematically by:
F = ma
Whereas in quantum mechanics, force is represented this way:
gradV([r]) = m(d2/dt2)
In the latter, the mathematical relationship remains, while what force ‘is’, as a thing, is replaced by a new formulation – gradV([r]). (Ladyman and Ross, Every Thing Must Go, 95.) Under certain conditions, the new mathematical equation gives the same predictions as the previous equation, in the same way that Newtonian equations remains useful when limited to strict conditions. It’s not the case, therefore, that previous theories are entirely dismantled and left to the dustbin of history. Rather their mathematical structure gets reformulated into a more precise and powerful (i.e. extensive) theory.
As another point against any claim of scientific dogmatism, we could also point to the philosophical literature. Schelling, Bergson and Deleuze all take science seriously, yet all of them offer dynamic theories of reality that are irreducible to a scientific theory that posits its own definitive entities as what truly exists. Moreover, science has moved towards dynamical theories as well – as people like Ilya Prigogine and Isabelle Stengers show, as well as the massive amount of work done by the Santa Fe Institute.
Which is all to say two things: one, the critics of realism who argue that is a conservative and dogmatic position are misunderstanding the nature of science. As the wild thought experiments involved in contemporary physics should make clear, it is far from conservative. And as the willingness to challenge any scientific truth should make clear, it is far from dogmatic. Science does not take its current theories as dogmatically true. The second point is that the critics who believe that realism must give a definitive answer and therefore that any internal debate is a sign that realism is false, are simply wrong. If debate were a sign that a theory was wrong, every single intellectual project of humankind would have to be considered false. Debate is not a sign of weakness, but rather a sign of vitality, and a sign of willingness to critique any philosophical or scientific dogma. Which is why speculative realism shares a common enemy in correlationism (and as the most intelligent articulator of correlationism, Kant). Speculative realism can very simply be seen as the questioning of continental philosophy’s long-standing dogma that holds correlationism as its unsurpassable limit. (And anyone who doesn’t see correlationism as the standard dogma in continental philosophy since Kant’s time need only read Lee Braver’s book, A Thing of This World: A History of Continental Anti-Realism, to see the progressive removal of all realism.)
Returning back to structural realism, though, James Ladyman and Don Ross’ book, Every Thing Must Go: Metaphysics Naturalized can be seen to offer a number of interesting ideas for any future realism. While the book itself is dense and filled with a number of insights (particularly as a representative of the explicitly scientistic position), two ideas in particular seem relevant to recent discussions.
The first has to do with the status of ‘objects’ in realism. The second has to do with the epistemological problem.
Now, as myself and Alex Williams have argued, the position of object-oriented philosophy (OOP) appears to entail a reliance on the phenomenological commonsense that reality is composed of objects. Yet it’s not clear (a) that physics requires objects in any meaningful sense, (b) that neuroscience doesn’t show that the brain constructs objects, and (c) that we can extrapolate from our phenomenology to what reality is like independent of thought. As people like Metzinger and Brassier have consistently argued, it is more than likely the case that any independent reality will be profoundly counter-intuitive. And as the mysteries and paradoxes of fundamental physics show, translating mathematical theories into intuitive parallels is fraught with competing interpretations, each trying to uphold a different aspect of commonsense.
What Ladyman and Ross add to this is twofold. On the one hand, they add the expertise to pronounce that there is a “convergence” in philosophy of physics towards the idea that there are no such things as individual entities in the study of fundamental physics. Undertaking an in-depth and extensive look at the various work being done in contemporary physics, they argue persuasively that individual things don’t exist. Rather, what exists are ‘real patterns’ – temporal and spatial patterns which are mapped by the mathematical structure of scientific theories. (ETMG, 120; also, cf. Daniel Dennett’s ‘Real Patterns‘ essay) Now patterns, in Dennett’s formulation, must be capable of being captured in a smaller amount of bits than the original data set from which they came. “The crucial point about Dennet’s version of this requirement is that he does not relativize what ‘could’ be computed to any special reference class of computers. Instead, he says, ‘a pattern exists in some data – is real – if there is a description of the data that is more efficient than the bit map, whether or not anyone can concoct it’. Thus there are presumably real patterns in lifeless parts of the universe that no actual observer will ever reach, and further real patterns whose data points are before our eyes right now, but which no computer we can instantiate or design will ever marshal the energy to compact. This is emphatically not instrumentalism.” (ETMG, 202-3.) In addition, real patterns must be ‘projectible’, meaning they must be capable of making accurate predictions about future events. Patterns that are incapable of sustaining predictions about unobserved events are not real patterns.
Yet, despite their antagonism towards objects within fundamental physics, Ladyman and Ross make the seemingly odd claim that objects exist in the special sciences (i.e. the sciences which model only a restricted portion of reality). Understanding how their ontic structural realism for physics can be nevertheless be combined with what they call ‘rainforest realism’ for the special sciences, will show the possibility of reconciling such a naturalistic position with OOP. In Ladyman and Ross’ work, physics doesn’t provide the fundamental level of reality to which everything is reducible, but rather the guiding framework within which other disciplines must operate. Or, in other words, the special sciences can have their own real autonomy, but only insofar as they are compatible with the findings of fundamental physics. How then, can individuals still be said to exist within the special sciences if current physics appears to prohibit them? The answer is in the nature of the real patterns – in some of the special sciences, real patterns will happen to take the form of individuals. It’s worth quoting Ladyman and Ross in full here:
“The main ontological implication of OSR is that reality is not a sum of concrete particulars. Rather, individual objects, events, and properties are devices used by observers (when these observers aren’t making mistakes) to keep cognitive books on what science finds to be sufficiently stable to be worth measuring over time, viz. some but not other patterns. Some of these patterns are indeed conceptualized as individuals in some special sciences, while simultaneously not being so conceptualized by other special sciences making projections at other scales of resolution. For examples, from the point of view of a behavioural economist measuring choices, a person is an extended pattern over a run of data, whereas at the scale where ‘common sense’ is most comfortable, but also at which social psychology is done, a person is the paradigm case of an individual.” (228-9)
Of course, the notion of keeping ‘cognitive books’ is likely to raise eyebrows amongst the anti-realists. But, (a) such patterns exist independently of any observer, (b) the notion of real patterns is determined in informational terms and is in principle available to computers capable of drawing natural laws out of raw data (which is already a reality), and (c) scientific realism’s fundamental claim is that science discovers real knowledge of a mind-independent world, i.e. it never denies that we use tools and ‘cognitive books’ to gain this knowledge. The metaphysical question here is whether this knowledge is justifiably realist or not, which again leads to the predictive success of science as a powerful indicator that it, in fact, does.
What this all means for OOP is that the notion of real patterns articulated by Dennett and extended by Ladyman and Ross is a useful conceptual tool for understanding the irreduction principle. It provides, in a computational manner, explicit criteria for formulating the reality of objects in the special sciences. But as Ladyman and Ross remind us, “one makes a metaphysical mistake if one reifies these essences and imagines that they are the real constituents from which the world is fashioned.” (241) In other words, it seems to me that while OOP can be a productive theory of ontic relations and potentiality, it may not be sustainable as an ontological theory.
Besides the notions of objects and real patterns, the second main idea of interest from the book is the shift in how observation is conceived. Adhering to a verificationist framework, Ladyman and Ross articulate observation not in terms of direct perception, but rather as being ‘informationally connected to’ – meaning there exists an information channel from the X under observation to the observer P. Note again, that in formulating it in terms of information, Ladyman and Ross avoid any anthropomorphic bias in regards to observation. It need not be a visual observation, and it is in principle open to an infinite number of variations of information connections (intriguingly, even as object-object connections, a la Graham’s notion of allure). While this notion of informational connection is sadly underdeveloped in the book, it can be seen as a possible retort to representational analyses, which continually invoke the unknowability of discerning between the representation and what is represented. Rather, in this formulation, any and all biases between the ostensible independent pattern and the observer are all incorporated into the informational network. So, for example, the theories which lead our attention to certain patterns are part and parcel of the informational connection; i.e. they are not a bias to be removed. In addition, as Ladyman and Ross state, “when we think about a relation of informational connectedness between some X and some P, we are thinking about both X and P as points (nodes) or regions (interconnected sets of nodes) in networks.” (308) In other words, this networked, informational form of epistemology negates any radical division between an observer and the observed. Or to put it in Laruellean terms, we are always already within the real. Both the observer and the observed are situated in an ontological framework underpinned by multi-scalar real patterns defined in non-anthropomorphic informational terms.
In principle, then, this type of realist ontology could easily be supplemented by cognitive science to begin to articulate how an information-based reading of the mind (what cog-sci focuses on) connects to the ‘external’ world in a consistent fashion. Finally, as Paul Humphrey’s work points to (cf. Extending Ourselves), all of these notions could be tied to advances in computational science, as well as AI studies, in order to begin thinking about the possibility of, in principle, unobservable and inhuman scientific discoveries made by computers. Speculative, sure. But the possibility is closer everyday.
[EDIT:] There’s also the not-so-small question of neuroenhancers, and their implications for shifts in cognition.
[UPDATE:] Alexei gives a thorough critique at his own blog. Asher adds some rather hilarious (in a good way!) thoughts into the mix. And Levi continues the debate at his blog.
Filed under: Uncategorized
There is a lot to respond to here, and I am doing the most annoying thing a blog commenter can do by pointing that out and then going on to say some rather simple and blunt things, so, yeah, apologies in advance for that. I think the sort of philosophy Ladyman and Ross practice is at distinct odds from the kind of non-philosophy as democracy of thought that Laruelle advocates. I won’t claim to speak authoritatively on L&R, as I’ve only skimmed the book, but it seems to me that this sort of scientism is at odds with both actual science and is philosophically suspect in that it clings desperately to a theological (Thomistic) problematic of erasing the human from any investigation into reality. The complete erasure of the human erases a part of reality and thus, to my mind at least, doesn’t get us anywhere nearer to the Real but occludes the very possibility of actually starting from the Real.
I am troubled, even as I respect it, with the obsessive turn to science (by which people always, always mean a certain strand of physics). Firstly, I should say that all of my work is a working with ecology as it currently stands (and thus should be open to revision as ecology progresses). So I’m not in any way anti-science. That said, it seems to me that most of the science worked with by philosophers interested in spec realism is knowledge gained from reading philosophers on science. That can tend towards an insidious folk conception of authority that is unacknowledged by all those folks who hate on folk knowledge (and, really, I feel like all this folk knowledge stuff has to deal with politics more directly and concretely w/r/t technocratic society). Meaning, without acknowledging, and Brassier is guilty of this I think, that you are working with a contested or concept that is still in debate (as he does with his cosmological model of heat death), you lend a kind of authority to the work that is not democratic (in the Laruellean sense).
Now I realize this puts me in a weird spot as the current doxa on blogs I actually like reading is that we should all be eliminativists now and that anything less is a kind of puerile idealism. Still, I think there are real live questions and that if this brand of realism is really to avoid becoming dogmatic it has to go beyond the kind of L&R scientism and to real, practiced science (even in the face of other demands to “shut up and calculate”).
Just want to add a thank you for such a considered and detailed post. I do hope my response isn’t taken outside of the spirit of intellectual friendship that I wrote it.
Thanks for the comments Anthony, and I think I should make it clear from the start that while I found Ladyman and Ross’ book extremely interesting, I don’t subscribe to all of it. My purposes were more to extract certain useful ideas from it that might be able to be transferred to another framework.
To respond to the direct critique about dogmatism though, I think Ladyman and Ross are very careful (much more so than others I’ve read), to be clear about the limits and openness of their approach. So, for example, they refuse to make claims about whether time exists or not (a contentious issue in physics), they refuse to grant authority to any of the competing replacements for QM and GR, and they refuse to take a stance on whether the 2nd law of thermodynamics is fundamental or derived. Moreover, speaking directly to the eliminativist question, their notion of real patterns is entirely supervenient on empirical research. So if science shows that mental patterns can be reduced to physical patterns, then we can be eliminativists about thought; but if not, then we have to follow the empirical results. All of those questions are open at the moment, and require empirical and theoretical evidence before one can take a stance on them. And to Ladyman and Ross’ credit, they are careful to only focus on what appears to have a consensus in modern physics.
While they don’t say so in the book, I’d disagree that they try to reduce the human out of the picture. In my opinion, the notion of informational connectedness entails that the human – with all our particular sensory modalities, as well as socially mediated theories – is an essential part in the information being transferred from the pattern to the observer.
As far as reading philosophers of science, rather than practicing scientists, I think that’s an important point. Certainly, there should be a balance between them at the very least. One (not so minor) hold back, of course, is that modern physics is a hugely complicated discipline in its own right! Most of us philosophers (Gabriel Catren being a notable exception) have to get our knowledge of physics indirectly – either through popularized accounts by actual physicists (Greene, Kaku, Hawkings, d’Espagnat, etc.), or through philosophers of science who specialize in physics. That being said, what I think you suggest your main problem is w/r/t to that is that it can appear that science is much more certain than it actually is. Which is why I think reading multiple sources and conflicting viewpoints is essential. I don’t know if you’d agree?
[...] 2009 April 21 by Alexei Nick has a nice post up at Speculative Realism discussing Ontic Structural Realism, which I was originally going to comment on, but decided that a full post is probably in order. I [...]
HI Nick, this is great stuff. One question: you write that talk of “cognitive books” will raise the ire of anti-realists, but surely its even more of a ‘red rag’ to the Eliminitavist position? I’m still unclear on how real patterns, though they can sometimes within certain fields be expressed as individuals/objects/discrete items, necessitate that these individuals are “real”… If they are simply a conceptual device, a useful mental handle on underlying patterns, how are they not simply an anthropic nicety which is ultimately still reducible to the real patterns (mathematisable spatial and temporal relationships) which they mark out? Im not sure how this vouchsafes the principle of irreduction at all…
And having just read Alexei’s critique: what stops this from becoming an exotic version of Platonism? Is it the autonomy of special sciences- but this seems to be pretty flimsy, the temptation will be to reduce back down to real patterns if possible?
“Moreover, speaking directly to the eliminativist question, their notion of real patterns is entirely supervenient on empirical research. So if science shows that mental patterns can be reduced to physical patterns, then we can be eliminativists about thought; but if not, then we have to follow the empirical results.”
Interesting, but doesn’t how one understands the data given by empirical findings determine what we think the data is telling us? To use (again, probably folk psychology) a Deleuzian phrase, doesn’t the image of thought one labours under determine how we understand the information outside that image?
For instance, the whole issue of “reduction” to me is very confusing largely because I labour under an expansive naturalist image of thought. What does it mean to say that we can reduce the mental to the physical?
“That being said, what I think you suggest your main problem is w/r/t to that is that it can appear that science is much more certain than it actually is. Which is why I think reading multiple sources and conflicting viewpoints is essential. I don’t know if you’d agree?”
I completely agree with you. It is why I read people, like Brassier and Metzinger, that, if correct, would destroy everything I work on. I suppose part of the ire that is raised, however, is a certain kind of Leiter-esque smugness towards whole swaths of philosophical practice. (Of course that is something that happens in every philosophical age though.) I mean, I for one am not so ready to consign Husserl and the whole of the phenomenological school to the flames even if I disagree with it. That isn’t to suggest that you’re doing that here, just to say that some of the “ire-raising” one sees amongst “anti-realists” may have something to do with this perceived smugness that may limit their actual engagement with the argument.
SBA,
Yeah, the eliminativists (and I’m still not sure where I stand with respect to it) would definitely see that as a major red flag as well. And L & R take an agnostic stance with regard to the truth of eliminativism. It’s up to empirical science to show whether it’s true or not. As I understand it, they would argue that the real patterns exist because they sustain viable predictions. Epiphenomena, on the other hand, don’t. (As for the question of subjective epiphenomena, Churchland has made the point that we are notoriously poor at predicting what people will do. And as Metzinger has argued, the sense of having made a decision emerges after the fact. So even though we may feel as though we make reliable predictions about others and ourselves, it’s not clear that that’s actually the case.) So real patterns can exist at multiple levels precisely because other levels are incapable of making the same predictions. One clear example that L & R cite is of a hungry tiger and a deer. To predict what the hungry tiger will do when it meets a deer, we wouldn’t look at them as a collection of cells and molecules. The real pattern only emerges from looking at it at a different level (and our observation of it doesn’t make it any less real). (p. 199)
As for the charge of Platonism, I tried to reply to it on Alexei’s blog: http://nowtimes.wordpress.com/2009/04/21/critique-of-structural-realism/#respond
Anthony,
“Interesting, but doesn’t how one understands the data given by empirical findings determine what we think the data is telling us?”
I think this is an important point, and something that L & R don’t give enough attention to. While I think they do, in principle, take into account the effects of theories on observations (through their notion of informational connectedness), there’s no real substantial discussion of this issue. (Which, to be fair, isn’t the main priority of their book anyways.) But I think it can be taken into account, with a lot of hard work. I don’t know though, it’s something I’m still thinking about.
“some of the “ire-raising” one sees amongst “anti-realists” may have something to do with this perceived smugness that may limit their actual engagement with the argument.”
I hope that isn’t the case amongst people who disagree with these positions! Levi’s position has consistently argued that all the typical correlationist concerns still play a major part. And while I lean more towards eliminativism than he does, I still think it’s difficult if not impossible, to not look at things like language, culture, subjectivity, consciousness, etc. There’s a ton of excellent material still available in ‘correlationist’ studies. I think the realist position is simply that correlationism has become limited, and realism wants to open the door to other types of analyses. And as I’ve said elsewhere, it would be a terrible mistake to confuse the enthusiasm we have for realism with certainty about our positions.
I like the hungry tiger argument. But though we don’t *need* to get down to the level of cells, atoms, strings etc to predict the behaviour of the tiger when it encounters a deer, to explain why it does so will lead us down a chain of explanations which head in that direction, would it not? So whilst from an occam’s razor PoV the simplest explanation, to get accurate predictive results, will not require elimination down to lower levels, the explanation OF the explanation will. Also,as regards the notion of patterns being real, but where it is also possible to explain a given pattern with a further one- so instead of objects all the way down, we now have patterns all the way down?
I think the intuition that any macro-level explanation will require micro-level explanation itself is something that Ladyman and Ross would argue isn’t supported by the current scientific findings – and so there’s no empirical reason (only metaphysical bias) to believe that it’s the case. The tiger example is one instance, but we could also think of traffic patterns in urban areas. It’s not clear that either will be reducible to some lower, purely physical basis. In that way too, the pattern is real and not merely an epistemic crutch.
Ladyman and Ross are fairly modest in their claims – they are hardcore scientists, and all their metaphysical speculation is based upon current science. Which also means its subject to change in the future. But it also means we need good metaphysical arguments if we’re to refute what current science says about things like reduction, or causality, or time. (Metaphysical arguments being something Ladyman and Ross themselves think is useless.)
Terrific post, Nick! I’m having trouble keeping up with all these books you’ve been referencing lately, but have been doing my best. It looks like I now have two more to order. A couple of points.
First, with respect to the different levels of strata from the quantum to the more developed, why can’t we treat these different strata as system specific emergent structures governing a system? Biology (and neurology) must conform to the laws of physics. I take that to be non-controversial. But when we talk about the phenomena investigated by biology, we are talking about something that is only possible based on the laws of physics (whatever those might turn out to be), but which cannot be deduced from the laws of physics. That is, the emergence of life requires very specific material conditions where matter develops its own patterns as a result of these conditions. Biology becomes the study of those particular patterned structures. These structures are not “epiphenomena” because they are entirely real and unique to that system, but they are also only possible by virtue of the laws of physics.
This brings me to a further point about maths and patterns. As per the Platonism charge, it seems to me that we must distinguish between maths and patterns. Patterns are real things that exist in nature. In my view, patterns are processes. It seems to me that every branch of science has more or less done away with objects, rejected the subject-predicate or substance-predicate logic that the concept of objects implies, and instead replaced these beings with events and processes. A process, is, of course, patterned. Maths, by contrast, are descriptions of patterns. It is not that maths are somehow causally “doing” something, but rather that maths are describing these mind independent patterns in the world.
I think this is a tremendously important point for a couple of reasons. First, take Jackson’s famous argument about the god-like scientist Mary who knows everything there is to know about physics, chemistry, biology, and neurology, but is brought up in a black and white room for her entire life. The famous question is whether Mary learns something new when she leaves the room and sees a rose for the very first time. We all agree, of course, that she does. The sophism of this little thought experiment lies in the idea that a description somehow produces the phenomena described. No one thinks there’s something mysterious about a hurricane that is there in addition to the description of a hurricane, nor does anyone believe that a description of a hurricane produces a hurricane or ought to be capable of producing a hurricane simply by being understood. I think that if we don’t distinguish mathematical description from pattern as such we fall into all sorts of spurious criticisms.
Additionally, I think great caution should be exercised with respect to treating maths as equivalent to primary qualities. In my view, the first chapter of Meillassoux’s After Finitude is a work of brilliance (though I’m finding very similar arguments all over the place now… for example, in Patricia Churchland’s Brain-Wise where she makes an argument against Kant from fossils), but I think his position, no matter how provocative and interesting, quickly goes downhill from there. He might be right that primary qualities are qualities that are mathematizable, but he seems to leap from this thesis to the conclusion that we can deduce qualities of being from mathematical thought alone. The problem is that maths allow us to formalize an infinite number of relations and patterns, but only a subset of these patterns are instantiated or present in nature. This is why we have to engage in the hard work of experimental investigation, creating information deriving technologies, etc., etc. I am unclear as to how we can conclude from certain properties of mathematics that the universe itself must be contingent. Of course, all the laws of nature are contingent in the sense that they could be otherwise and that there is no logical necessity that they be thus and so. But it is not at all clear to me that they are contingent in such a way that they could change at any point in time.
Finally, with respect to Anthony’s remarks, it seems to me that realists aren’t so much “scientistic” as they wish to introduce science into philosophical discussions. I think one of the problems here is rhetorical. In order to address any problem– in this case the absence of science in continental philosophy –you have to discuss that thing. This gives the reader the impression that this is the only thing that the person is concerned about, when that isn’t the case at all. The situation is not all that different from debates surrounding feminism or gay rights. The detractor declares that women and homosexuals only care about women and homosexuals and want to subordinate everyone else to their “agenda”, when they’re just looking for an equal shake. Truth be told, I think it’s deeply scandalous that science is all but absent in continental orientations of thought. How is it possible that we live in an age following the Galilean revolution, where we’ve seen massive developments in physics, biology, neurology, etc., and these things are blithely ignored by most continental philosophers? How is it possible to ignore these things when they so fundamentally shake the foundations of key assumptions in the history of philosophy?
As for the issue of the place of man, I don’t see realism excluding the human at all. Rather, I see realism as pointing out that all things aren’t correlated to the human, that they aren’t all related to the human, and that there is a real that is in no way dependent on the human. In no way should this be taken as equivalent to the claim that the human should be ignored or that it is irrelevant.
Levi, I have to be honest, I really don’t want to have a protracted discussion with you as things often go downhill with you and I. I’m sure that has something to do with the laws of physics or something. I do want to point out, with regard to my observation about “realist smugness”, maybe a more careful analogy could be thought of than one that essentially puts those who disagree with the realist position in the same camps as homophobes and misogynists.
Levi,
I agree with what you say about the emergence of real patterns, although Ladyman and Ross are resistant to any notion of different levels of reality. Instead of hierarchical levels, they argue more in terms of scope – whereas physics is the science capable of providing data from any measurement in the universe, the special sciences are concerned with patterns discerned under restricted conditions (e.g. mid-level phenomenology).
I think you’re right with your response to Platonism as well. While the mathematical descriptions change as theories change, the patterns themselves stay the same (since this is what accounts for the success of science).
The distinction between logical and physical possibility/necessity is, I think, increasingly important in these discussions. L & R point to physical necessity as their basis for maintaining the empiricist credentials, as well as for refuting Platonism. Yet, what does something like the ‘many worlds’ interpretation of QM say in this regards? Is physical possibility too constrained, thus opening up the possibility of mathematical ontologies? I don’t have an answer, but I can see similar questions becoming a big dividing line between people like Meillassoux and Brassier.
“The situation is not all that different from debates surrounding feminism or gay rights. The detractor declares that women and homosexuals only care about women and homosexuals and want to subordinate everyone else to their “agenda”, when they’re just looking for an equal shake.”
I think, Anthony’s caveat aside, that this is a pretty apt metaphor. The absolute neglect of science within continental philosophy is brutal (Deleuze being a notable exception). At the very least, to start having real discussions about science in continental philosophy would be a great step forward – without the preemptive attempts to justify ignoring its findings. (Usually by trying to reduce science to being a derivative of some pre-scientific level of experience amenable only to philosophy.)
Nick,
An interesting post with a good reading of Ladyman and Ross. However I’m still not convinced by your repeated claim that science has been neglected by ‘continental philosophy’, because it seems more accurate to recognize that it is just the texts you are choosing to reference. Perhaps If you wish to discuss science in these cross philosophical disciplinary terms, then explicitly incorporating analytical and continental philosophical traditions might be more useful than continuing to critique a non-existant entity called ‘continental philosophy’. The continental philosophy I’ve existed within has always contained philosophers such as Bachelard, Canguilhem and Foucault with a strong relationship to science, that was known long before I became aware of Latour and Stengers work in the 90s, just as the analytical tradition contained mathematical realists and naturalists such as Penelope Maddy (whose critique of Kant predates Meillassoux critique of Correlationism by some years… )
The problem is that I don’t see how the argument that Continental philosophy brutally neglects science is sustainable, but I can see how it is in your ideological interest to claim that it does, in the furtherance of your emerging position. I think you need to be more honest, open and extend your references deeper into the analytical tradition.
It probably won’t convince me, just as Maddy’s critique and refusal of Frassen and by implication Deleuze does not convince me but I am not your target audience anyway…
best wishes
steve
Steve,
I think you’re definitely right here. Part of my claim is certainly polemical, and I should be careful to more accurately state it. My claim about the neglect of science is more related to my own personal history, and to the English world’s appropriation of French philosophy. Canguilhem and Bachelard were certainly much bigger figures for the original French thinkers than they ever were for their English derivatives. So when I say there’s been a neglect of science it more accurately refers to people working under the influence of Heidegger, Derrida, Nancy, Lacan, and Foucault (who did focus on psychiatry and social sciences, but never the natural sciences).
I think this situation is changing, as Latour, Stengers, Malabou, etc. all show – but in my own experience (which is hardly generalizable), these are still far from being dominant or normal positions.
And I’m definitely attempting to expand my references into analytic philosophy (as I think this post reveals), but along with expanding my reading into neuroscience, cog-sci, physics, and German idealism, I’m running out of time! : )
I don’t disagree with this, but it demonstrates the significant difference between my personal history relating to Continental Philosophy, which could be considered as being marked by science and engineering to a much greater extent than yours appears to have been. With the consequence that I have a less romantic understanding of science and scientific realism than you appear to. I have always been closer to the lines of thought bordered by engineering and science than the linguistic turn that is evident not simply in the Heidegger and Derrida lines but also in the Analytical lines of Wittgenstein, Rorty and co. My liking of Deleuze is founded on his constructivism and the three critical terms of engineering, machines and the material – an engineer will always be a constructivist and consequently a non-realist at least as we have discussed it elsewhere. Where a scientific realist will always tend towards believing as Kostelecky says “we might just catch a glimpse of the ultimate theory that underpins our universe” wheras an anti-realist, a constructivist will always veer towards the more limited and interesting case of the phenomenological laws that can save the phenomena and have supporting evidence. Either way whilst Kostelecky’s realist sentiments may be a problem for a philosophical realist or naturalist, they are certainly not for an anti-realist as until the phenomena can be proven we must regard the fundamental theory as unproven, less interesting than the more useful and frequently incorrect theories we work with everyday. A further difference is that you have suggested that science is counter-intuitive, this may be so to a realist but it is certainly not the case here – it may be problematic and appear counter-intuitive to a realist that the theory of relativity and the standard model of quantum mechanics have incompatible concepts of time, but an engineer and an anti-realist always works with incompatible concepts that sometimes have thin supporting evidence, but because we work with phenomenological laws rather than fundamental laws this is not a significant problem. A realist must be bothered by the problem of time mentioned here because it throws into question the fundamental laws which support their realism. If such a unified law can be proven then we would of course accept it, but this is not the case for all scientific laws there may be good political reasons why it needs to be rejected.
For most people engineering remains the technical application of the results of science, personally I think that engineering and philosophy are directly related. They are equivalent in there anti-realist and constructivist directions… The difference it seems to me is that you want fundamental laws to support your realism, whereas anti-realism doesn’t mind the incoherence between multiple laws as long as the phenomenological laws agree with the evidence, and save the phenomena !
Incidenatly I think that you can best understand Foucault’s writings on the human sciences if you recognize his understanding is founded on a particular reading of the natural sciences….
nick-
as a first time reader of your blog from the left coast of the us, I am taken by the quality [pardon the pun] of discussion and discussants. I will be happily following from afar. coming from a heideggerian/merleau-pontian POV I find this all fascinating.
sdv-
“The difference it seems to me is that you want fundamental laws to support your realism, whereas anti-realism doesn’t mind the incoherence between multiple laws as long as the phenomenological laws agree with the evidence, and save the phenomena !”
as an architect I have to agree. we must process all kinds of ontic information yet creativity comes from not denying that facticity but going beyond it to a kind of recombinative changed relations- new phenomena not always understood-but with discernible ’something’ that the idea of pattern and field noted above does not seem to negate, or for us, is unnecessary to substantiate. the thing as pattern v object description I find totally agreeable and places it within flow and process and various notions of temporality. this is especially important in larger scale environmental interactivity of things and others-a knowing and thinking where the mathematical basis of alternate or foundational description seems way behind, but emerging.
I am fascinated that husserl is not noted more, as his ’science’ of pheneomenology project[s] and various forms of reductions may hold some valuable tangent possibilities.
pardon my inadequate word choices, I usually draw.
Thanks talkingbrick, it’s always good to get new voices involved!
Sorry, Anthony, I honestly didn’t in any way mean to suggest that you’re a sexist or a homophobe. My remark might better be understood in the context of these discussions. Since we’ve begun discussing these things online I’ve had people shoot things at me like “you just love science”, “you have a love of all things scientific”, “you just hate Kant”, “you’re a vulgar positivist”, “okay we get it, you’re a realist, why do you have to keep talking about it?”, etc. To me these charges have had the flavor of the straight guy saying “I have nothing against homosexuals but I don’t see why they need to talk about it all the time.” I find these sorts of remarks especially odd because it’s pretty clear that I work in a variety of registers, many of which are correlationist, ranging from the phenomenological and the semiotic to more basic metaphysical claims. In my own forays into science and realism I’m certainly not suggesting that science should replace philosophy or that all philosophical questions can be answered through science.
All of that aside, I just don’t sense this smugness you’re attributing to those who find value in science in philosophy. In my view, scientific claims are pretty humble. They are proposed in the spirit of tentative finds or most likely conclusions and they invite public criticism and critique. If anything strikes me as smug it is the implicit continental attitude that science is a priori dogmatic. On the one hand, I think much of this has to do with how continental thought evolved from the Romantic movement which was hostile to science and mathematics. On the other hand, I think Heidegger was a big influence here, convincing many continentalists that science is a synonym for the metaphysics of presence. Truth be told, I get the sense that a lot of continentalists think that scientists are a bunch of dense, superficial rubes and that continentalists have a sort of secret or esoteric knowledge that no one else possesses.
“Truth be told, I get the sense that a lot of continentalists think that scientists are a bunch of dense, superficial rubes and that continentalists have a sort of secret or esoteric knowledge that no one else possesses.”
This is actually a perfect example of the smugness I’m talking about. I did my undergraduate degree at DePaul, host to some pretty hardcore Heideggerians. Yet, aside from the occasional quip by a person here or there, did I ever hear talk of the worthlessness of science. Naas, most well known for his work on Derrida but whose main passion is ancient philosophy, once tried to convince a student that math gave access to eternal truths (there was some subtlety to it I’m sure as this was 4 years ago), would go on about the love of Einstein in Joyce’s fiction. Bill Martin, a Marxist philosopher, also would often go on about the value of science even while distrusting a kind of scientism present in the CPUSA (And, to Nick, if Continental philosophy has been so disenchanted with science how do you explain the predominance of Marxist tropes in Continental philosophy? Tropes that always looked to the social sciences.). Peter Steeves, one of the best practitioners of Husserlian phenomenology I can think of, worked for NASA and makes constant reference to science in his written work. An ecologist friend of mine at DePaul currently doing a PhD in philosophy has noted this acceptance of science as well and he has a very different complaint than the (in my view erroneous) opinions I see espoused all the time by yourself. For him philosophers need to become more bold in their working with science and encourage a richer interaction, but he by no means sees this assumed “esoteric knowledge”. This is not to mention the work of Babitch, Tymieniecka, Protevi, Zahavi, some Merleau-Ponty-ists (to just mention American Continentalists). It is just false and arrogant to think that this is the first time Continental philosophy has taken science seriously and, further more, it is distracting from the actual debate.
And I’m not anti-realist, not in the least, I just don’t know that real things have to be known independent of the human and I’m not sure I understand what that even really means.
Just to quickly point out that Sokal and Bricmont wrote their little book not because contemporary French philosophy, Continental philosophy”, and theory didn’t do enough with science, but because they didn’t think they should be talking about science at all!
[...] the latter), though placebos are a lot more interesting than we ever imagined (via Nick at Speculative Heresy). The point is that the normative does not tell us what is and is not. Rather, if we are to [...]
I definitely wouldn’t disagree that there’s a lot of great continental work that incorporates science into it. But in my own personal experience (outside of the blogsophere, and outside of impersonal articles I hunt down – i.e. outside of areas where I’m prone to meet like-minded individuals), the reaction to scientific claims has almost always been disparagement by continental philosophers. It sounds like you’ve been lucky enough to mostly avoid that – and I do think that there’s a significant change occurring w/r/t to continental philosophy’s relation to science. (With a simultaneous shift in analytic philosophy’s relation to continental philosophy. Our generation, in my opinion, tends to see the divide as arbitrary.)
As far as Marxism, I think the references to ’science’ are much more rhetorical than anything else. I mean, in some respect their following in the lineage of Hegel’s Science of Logic, and not natural science. So they can claim to have uncovered the objective structural dynamics of society, yet these aren’t discovered in the same way that, say, cognitive science would uncover mental dynamics today. But the perceived authority of science has and always will have a powerful attraction. Just read any works of social science today and you blatantly see the desire to be ’scientific’ – which basically just entails using math.
Anyways, this is all to say that debating about whether continental philosophers do talk about science is a matter of tendencies, and not absolute facts. It’s also a matter of personal history, and I know Levi has written elsewhere about the philosophical environment he grew up in. And to add to that, I completely agree that it’s distracting from the actual debate!
I’m not sure how I’m being smug, Anthony. I applaud the fact that all of these professors you’ve studied with are celebrating science, but I can’t say I’ve read a single page of Naas’s work where discussions of science are front and center. Moreover, any causal perusal of the proceedings for SPEP will show that references to the sciences are almost entirely absent in the papers presented there. The same thing can be discerned in Continental publications in journal and book form. Merleau-Ponty is a genuine hero here because of how he took science and psychology seriously, but he’s the exception rather than the rule. For every anecdote you put forward about a Continentalist defending science I can provide another for a Continentalist dismissing it.
I think the anecdotes you put forward indicate another dimension of the problem. You have Naas eulogizing about mathematics and Bill Martin (great guy) going on about the value of science. But where are they doing it? In the classroom and in discussions. Why isn’t this present in texts? There seems to be a strange sort of split among Continentalists where outside of our published work we are big fans of physics, biology, neurology, etc., yet this seldom directly becomes fodder for philosophical thought in texts. It’s as if all of that has to be bracketed when doing philosophy. I think this is directly related to the two major trends of Continental thought: linguistic idealism and phenomenology. The former deals only with texts and signs, and therefore has no plausible way of introducing science into its framework. The latter deals with the given and givenness, bracketing the natural world, and therefore has no plausible way of introducing science into its discussions in any but negative terms as part of what’s bracketed in phenomenological analysis is causality which is the bread and butter of all scientific investigation. Again, Merleau-Ponty is a notable exception here and things have begun to change somewhat (K.A. Pearson, DeLanda, Badiou, Brassier, and so on), but these are still very much minority positions.
You write:
I think this is a distortion of the SR thesis. The SR thesis is not that things “have to be known independent of the human”, but rather that the real is a world independent of humans. Clearly humans must relate to the world to know that world, but this relation doesn’t “make” the real what it is.
I disagree with your appraisal of Sokal and Bricmont. What they were railing against was anti-realist claims about science that reduced it to one more “postmodern construction”, not claiming that philosophers shouldn’t talk about science at all. They believed that postmodernists were trying to undermine the truth of science and tried to discredit them by revealing that they didn’t know what the hell they were talking about. This is, of course, a highly inaccurate reading of what postmodernists were doing with science.
[...] Alexei at Now-Times wrote a post about a post by Nick at Speculative Heresy, which, thankfully, was not about a third post, but instead was about Ontic Structural Realism, [...]
Levi,
Really now, we should just stop. I want to say one final thing and then the more interesting and important conversation should go on.
The point of the anecdotes was not to show that Naas or Martin (though Martin does engage some with science in Ethical Marxism) make science front and centre in their work. It was to point out an obsession you have with trying to show that your work and the work of others who are taking a more intentional stance towards science is somehow a radical break with Continentalism. It simply isn’t. And, yes, of course you could counter any anecdotal evidence with your own. But for every analytic philosopher who is careful in their approach to science, I can show you a dogmatist as well. The field is – my God! – varied!
Furthermore, it is hardly the case, to my mind, that engaging with science automatically makes your philosophy more interesting and worthwhile than, say, John Mullarkey’s recent book on film and philosophy, which you again seem to often suggest.
I suspect most philosophers, analytics included, are weary of engaging too directly with science because they may get it wrong. I mean, have you talked to any physicists about your work?
Levi,
Sokal and Bricmont were actually doing a combination of your interpretation and Anthony’s.
1. Anti-realism undermines science, and this is a bad thing. Which is your interpretation, broadly.
2. That pomos didn’t understand the science they were using their philosophy. Hence they shouldn’t use science. Anthony’s interpretation.
Funny thing is, over at Camp Analytic (purely an ideal place in my mind) they are probably cackling like old crones – and now the continentalists agree that what they were doing in the past was anti-science! Remember the theory wars, well we have won the ground!
These are some wonderful discussions here everyone. It makes me overjoyed that the essays on science by Laruelle fits so snuggly into all these nooks and crannies you guys have been hollowing out here. I don’t really even have time to expand on the concept (at the moment), but I hope to be able to reply later with some thoughts on the “generic” sciences, and what that means for L. One of Laruelle’s recurring claims is that in order for democracy (not a political position but the equality of disciplines) to enter into thought, there must be a peace treaty from philosophy to science which would no longer put philosophy above science in a hierarchy. That’s only the theoretical basis for the ‘generic’ though, and doesn’t really expand upon that concept. Save that for later.
Anthony,
I think those ‘laws of physics’ you spoke of earlier must be kicking in! And we all know you can’t fight gravity.
It’s just a sidenote, but maybe worth it’s worth mentioning nonetheless: there are about a billion different kinds of realism, very few of which entail one another (I still am curious to find out, for example why ontic structural realism doesn’t collapse into epistemic structural realism, and hence into a version of Kantian anti-realism/transcendental idealism). So one is perfectly able to be a realist, say, about certain kinds of unobservables (E.g. WIMPs) and an anti-realist about others (e.g. loops); moreover one can be a realist about mathematical entites and an anti-realist about unobservables; one can be a realist about meaning, but an anti-realist about moral notions. Following Putnam, one can be an internal realist without having to be an external, metaphysical realist. The list goes on — and on and on… It’s almost tedious.
All this to say, I suppose, that (1) one can’t simply use the label ‘realism’ as if it meant a determinate, well demarcated thing; there may be a something like family resemblance amongst all realisms — it may even be productive to try and articulate it — but there’s no non-contentious definition of ‘realism,’ past the truly uninformative one concerning the independent existence of something or other. And (2) since there are so many species of realism one really has to motivate one’s discussion for a particular kind of realism, by (2a) showing what rational advantage(s) this commitment purchases, (2b) showing that this brand of realism uniquely confers these advantages (2c) that this form of realism is non-contradictory, etc.
Personally, I tend to think that most realist/anti-realist debates are red herrings. Science gets on just fine these days without philosopher’s puzzling over the ‘reality’ of this or that thingie, which no one really understands yet anyway. To echo Anthony, there, there are so many different, groundbreaking, and interesting approaches to various philosophical issues. Why should everyone all of a sudden decide that ‘realism’ is the most problematic notion for everything under the sun?
From my own experience with various forms of realism, the typical answer runs something like this (and I think this is Levi’s motivation too): Without something mind independent, there is no objective grounds for the truth of our claims about x, y, or z. That’s fine I suppose. But it was this kind of thinking that made it impossible for an atheist to act morally in the 18th and 19th Centuries; it’s why people still believe in God (I heard some dude on TV claim that without God, there’s no objectivity in, or value to our actions — i.e. everything goes); it’s why folks in the USA still swear on the Bible (or other religious book) to tell the truth and nothing but the truth in courts of law(!). Point being: ‘realism’ — no less than anti-realism — play a justificatory and explanatory function within a philosophically/historically (take your pick, it doesn’t really matter at the moment) constructed space of reasons. For those of us interested in the contours of this space, the various forms of realism aren’t particularly helpful or interesting in and of themselves.
But I should leave off here, since I think I may be rambling.
Alexei,
I’m headed out the door, so just a short reply for now. But you’re certainly right that there’s a ton of different varieties of realism, and so the generic term itself loses a lot of value. (Which is also why it’s important to distinguish between the varieties when we’re talking about the justification and the implications of realism.)
I think, in response to the value of the realist question itself though, that any of the speculative realist works show why it’s an interesting and still important question. Namely, things like the necessity of contingency (Meillassoux), assymetric causation (Harman), and unilateralization (Brassier) are all interesting insights that are only possible in a realist framework. That’s not to say that they’ve been proven ‘right’, but at the very least they expand the scope of philosophical imagination and open up new paths for thought to take. Speculative realism, as opposed to scientific realism, is a much more interesting venture that way – rather than seeking to ontologically justify science’s claims (which, as you rightly note, most scientists don’t care about anyways), speculative realism tries to understand the logical and metaphysical implications of realism. And I still think something like Brassier/Churchland/Metzinger’s eliminativism has profound implications for ourselves.
Alexei,
I think you’re on mark with all the different variants of realism and anti-realism. For example, there are a number of places where I remain a correlationist. You write:
Perhaps. However, when you write
I think this misconstrues the issue somewhat. To my thinking, realist philosophy isn’t addressed to the scientists, but is a problem or issue internal to philosophy. I fully agree that science gets along just fine without these metaphysical and epistemological debates. If the realism/anti-realism debate is to have significance it must lie elsewhere.
I am not entirely opposed to treating the “real” as a sort of regulative ideal we’re constantly working towards. This is one reason that I say things like “occasionally we grasp a bit of the real” (a Lacanian mode of expression), because I take it that the real isn’t something we have, but something we’re perpetually working towards.
However, suppose we take a pragmatic justification for adopting some variant of realism. Here the question is, “what are the consequences of adopting an anti-realist position?” I think that if we look at the subsequent history of Continental philosophy since Kant we see that the anti-realist position has had a number of negative effects. By and large, Continental thought has come to be dominated by the analysis of texts and engagement with the history of philosophy. While living in the midst of massive technological, social, and scientific revolutions that fundamentally challenge our understanding of the world, these matters have gone largely unremarked. Sure, there are exceptions here and there, but they tend to be at the margins or periphery, or when they are not (Heidegger talking about science, maths, and technology) they tend to be rather reactionary.
I think this absence can directly be traced to the rise of correlationism as the predominant position within Continental thought. Let’s take the two “big boys” of Continental philosophy: Phenomenology and the Linguistic Turn. Both of these approaches to philosophy have an internal logic that renders them ill equipped for addressing these sorts of issues. Insofar as phenomenology brackets the world so as to get at the pure givens and givenness, it necessarily ends up bracketing all of science because it has to bracket causality. As such, from a phenomenological perspective, scientific claims come to appear inherently dogmatic or as claims made from within the natural attitude. Likewise, the Linguistic Turn effects a reduction similar to that of the epoche of phenomenology. Here the reduction is to texts, language, signifiers, signs and how they construct their objects. As a consequence of the Linguistic Turn we get a sort of “de-real-ization” of referents, excluding them from philosophical discussion a priori. To speak of a “real” referent from within linguistic forms of correlationism is to fall prey to a naive and dogmatic illusion where we believe we can talk about real referents rather than linguistic constructions. Again, science disappears or is reduced to one text among many. Not surprisingly, since the Linguistic Turn has concluded that everything is text, it contents itself with the endless analysis of the texts of the tradition, suspending any evaluations of the claims of these texts in terms of truth.
From a purely pragmatic point of view, this is a scandalous situation. Here we have these massive transformations taking place, yet philosophy has little or nothing to say about these transformations. Yet isn’t one of the central vocations of philosophy to reflect on its time, especially when that “time is out of joint”. It is not by mistake that philosophy takes place or occurs during periods of political, social, or scientific revolution. Contemporary philosophy, I believe, has failed terrifically in this respect. Were we to adopt realism as a regulative ideal we would be less susceptible to these bracketings and their deleterious consequences.
A few rejoinders:
Nick — have a good day ‘outdoors;’ I suppose I’m simply not that impressed by the SR work I’ve read (which is far form exhaustive; I’ve read QM’s little book and skimmed Harman’s two, but I haven’t been able to get a hold of Brassier’s latest or Grant’s stuff. I haven’t read Laruelle.) I simply don’t think there’s anything novel about QM’s emphasis on contingency (Rorty said much the same thing at least 20 years ago in Contingency, Irony, Solidarity), or in his critique of certain forms of correlationism (again, Rorty talks about this in terms of ‘mirroring;’ see Philosophy and the Mirror of Nature). I don’t see why Harman’s ‘objects’ aren’t either Kant’s things in themselves about which we can reason but not know, or a normative structure akin to a Husserlian formal ontology with its correlative formal logic (with the transcendental logic simply excised, or inverted so as to ground his analysis of physical relations rather than ‘real’ ones). However refreshing the arguments are, however revelatory they may be for died in the wool ‘continentalists,’ they aren’t groundbreaking. The only thing that is novel in QM and Harman is an explicit return to metaphysics, which seems to mark the end of post-metaphysical thinking — indeed in much the same way as the dreaded ‘religious turn’ marks the end of ’secularism.’
My simple point is that although the explicitly metaphysical-realist interpretation is ‘new’, the SR folks are, to use one of Blumenberg’s notions, simply ‘re-occupying’ an older (largely anti-realist) problem space, and thereby shifting its emphasis. My question is simply, What accounts for the reoccupation? What is the precise failure that leads to the return to metaphysics — indeed to the re-ontologization of post-metaphysical philosophy? I mean really, if you can generate the same results (i.e. the necessity of contingency, and the critique of correlationism) from both a realist and an anti-realist position, how do you choose? Moreover, given that most of the work of Brandom and McDowell is a response to Rorty’s challenges, why isn’t SR not looking at their work as a resolving their issues already?
Levi — I agree that the significance of ‘realism’ lies elsewhere than in actual scientific activity. I said as much already. IN fact, I said it lies precisely within the sphere of giving and taking reasons. This said, I simply disagree with everything you said thereafter. I don’t think anti-realisms are pernicious in the way you suggest (they’re certainly no more pernicious than the various forms of realism out there), nor do I think that the lingusitic turn does what you say it does (Austin was a direct realist, see his Sense and Sensibilia, Quine was a naturalist, Searle is a realist, before his pragmatic turn so was Rorty, and the list goes on), nor do I tend to think that the various phenomenologies ignore the maths and sciences, so much as they propose a different, incommensurable field of investigation (i.e. some form of explanatory gap). The being-sense of the world is closer to us that the transcendent world. It would take too long to explain why I think this, so let’s leave it at that. There is a fundamental difference in perspective between us, a differend I suppose, which I don’t think will be resolved in this discussion. Pace.
Alexei!
Read what I said carefully! Throughout I referred to problems within the domain of Continental philosophy, so the reference to Quine, Austin, Searle, etc., is not relevant. Anti-Realism has certainly led to all of these problems institutionally within Continental philosophy. I don’t see how this can possibly be denied when examining the publications in Continental journals, from Continental presses, and at Continental conferences.
I stand corrected then.
Levi, you keep giving these very basic descriptions that I have already contested many times, of course, it’s up to you to keep giving them, but really your description of SR is just a description of any realism – there’s a world out there, we can relate to it, but in relating to it we do not create it – that’s naive realism that puts all the issues we covered so many time (epistemology, access to the world, observer’s prejudices etc etc) aside and proceeds as if nothing happened…
I actually have this Ladyman/Ross book and I was eyeing it suspiciously for a while now – honestly now, Nick, is it going to make me cry? I’d like to take a look at it after I read your post – most of it too complex for me anyway, I like my philosophy like I like my Kant – boring…
[...] 24, 2009 by Mikhail Emelianov Unfortunately for me, I’ve missed a post on Speculative Heresy from some days ago on realism – with some remarks there and on his own blog, Alexei responds to the [...]
How can the decision to accept eliminativism be empirical? Eliminativism ultimately rests on a decision, a moment of freedom not able to be encompassed in its own account. Empirical science can show us the usefulness of a certain type of eliminativism, but its truth or non-truth is not an empirical question.
Mikhail,
It’s a great book, but it does bring in a lot of disciplines I wasn’t very familiar with myself. So it’s difficult, but they write clearly and explain alternative ideas, which means you have that contrast to make sense of their own position.
Frank,
I think your position relies already on the non-eliminativist assumption that there is a moment of freedom. Take away that assumption and it becomes an empirical matter. We may not be able to experience the eliminativist position from a 1st-person perspective (by definition), but if science shows that we can produce illusions of control and free will in people, that’s a major stumbling block to our 1st-person illusions. I have little doubt that we could go on believing otherwise, but that doesn’t make it not true. It would be a type of knowledge that we would abstractly know (I have no free will), but would never be capable of experiencing (because of the way we, as subjects, are constituted).
I’m thinking about the importance of identifying patterns in the “special sciences.” There’s a statistical technique called Factor Analysis that automatically identifies patterns in raw data independent of theory. Say you’ve collected multiple data points on multiple subjects. You construct a big matrix of all data on all subjects, then have the factor-analytic algorithm look for statistical correlations within the matrix. The algorithm iterates through all sorts of possible formulae for combining and weighting variables, then it spits out a set of equations — the “factors” — that together explain the greatest proportion of variation throughout the data matrix. You can even specify the number of factors to be included in the solution.
The “art of science” consists primarily of making theoretical sense of these factors. You tweak the number of factors, maybe you lock one or two into the model and have the algorithm explore the remaining variance, and so on. You’re looking for some combination of statistical power, elegance in the model, theoretical comprehensibility, and perhaps other more pragmatic considerations. Eventually the scientific community converges on what it regards as the best solution, at least for the time being. E.g., the “Big Five” model of personality is the currently-accepted standard in the field today, but no one regards this 5-factor model as “real” or permanent.
The point I’m trying to make is that the factor analysis algorithm can generate a vast number of solutions by identifying pattern information embedded in the data. Each factor-analytic solution is statistically valid; it’s even possible that each solution could be explained by some theory. There are big interrelationships between raw data, patterns, mathematical analytical techniques, and explanatory concepts. All of them come into play. I suspect the same principles apply in the more basic sciences. E.g., I’m guessing here, but I suspect that whether 10-, or 26-dimensional string theory ultimately comes to dominate the field will probably depend on some combination of elegance, predictive empirical power (if that ever becomes possible), and theoretical justification.
Oh, and one more thing. Yesterday I happened to see a poster display describing a study in scientific education. The researchers have ascertained that university students taking intro chem or physics classes tend to regress during the semester to more novice-level attitudes toward these disciplines. One indicator distinguishing novices from more advanced scientists is the extent to which one believes that scientific concepts can be understood without knowing the math. Novices said no, that the math is essential to understanding the ideas; the experts said that understanding the ideas can be achieved without knowing the math. I.e., expert scientists regard language as a legitimate scientific tool.
Really interesting stuff John, even if it brought back bad memories of undergrad statistics!
I’d be very interested to know why experts think the math can be left aside though. It seems to me that one of the main problems in quantum physics is that the math gives a straightforward answer, while translating it into common sense is far from clear. I suppose it depends on what is meant by “experts said that understanding the ideas can be achieved without knowing the math”. Certainly, as any number of popular science books make clear, one can communicate and understand these theories without having to rely on math, but at the experimental level, it still seems necessary to me. (But maybe that’s just my novice understanding?)
As far as patterns go, I think psychology is in a bit of a unique situation since it’s not really a mature science in the way physics or chemistry is. But I do think your point is important – and it highlights to me, one of the main drawbacks of Ladyman and Ross’ book. Namely they point to the institutional structure of science as guaranteeing that ‘real patterns’ get accepted. And maybe this idea works well in physics? But as you point out, it’s far from clear that the institutional mechanisms are sufficient in psychology.
My main point from the factor-analytic example is that whole swarms of patterns can be identified within phenomena, yet these patterns might not mean anything in and of themselves. Patterns typically are not readily evident by eyeball inspection; sophisticated mathematical algorithms are required to extract them from the data — a situation not unlike that found in the hard sciences. But the patterns are there in the phenomena themselves, discovered through math rather than created by minds. Does this mean that the real is inherently mathematical? Not necessarily.
Psychologists used to fantasize that, as the field matured, the inelegance and divergence of its theories could be reduced to relatively small numbers of underlying variables and formulae linking them together. This reduction is successful only by artificially limiting the complexity of the phenomena under study. Assuming that humans are composed of a relatively small number of elements assembled together through a small number of basic forces is to reduce science itself to some universal standard of what truth ought to look like. I would expect an object-oriented theory of science to acknowledge that the complexity of a discipline corresponds to the complexity inherent in the world of objects under investigation.
I’ve not seen the whole study, but I’d say that working scientists regard math as a tool for digging into the world, not as something which corresponds directly with the world. Math adds precision to observation and to the formulation of theories. So too does linguistic precision. Math is itself a language, wouldn’t you say, with its vocabulary, grammar, and syntax. Verbal and mathematical languages can both be used to describe the real without necessarily representing the real. Dismissing verbal language as a sort of barrier blocking humans from direct contact with the real is, I think, an artifact of certain strains within continental philosophy.
Just because something can be stated verbally doesn’t necessarily make its meaning transparent to common sense. Science is known as much for its jargon as for its math. It’s not unlike expert philosophy in that regard, where the language is off-putting to novices and often dismissed as mere gibberish.
Hi Mikhail,
As I’ve outlined elsewhere, this is not what I understand by “naive realism”. Were this thesis “naive realism” then every version of realism would be naive. I’m perplexed by your final sentence. The realist does not ignore epistemology, questions of access to the world, or observer prejudices. Indeed, the realist is especially concerned with the latter, which is why the scientist goes to so much trouble to create controlled experiments. The recognition that there are observer prejudices is not equivalent to the correlationist thesis that we only ever know phenomena. Again, I invite you to read Roy Bhaskar’s Realist Theory of Science for an account of realist inquiry that addresses these sorts of criticisms.
Levi, all I was saying that you seem to be repeating a number of definitions of realism only then to qualify them if they are questioned in any way, yet even if you agree to some objections and there seems to be a progress in our discussions, you then go right back to the original formulation. Recall the discussion of primary and secondary qualities – I was very impressed with the fact that you have honestly admitted that it’s still very much in the works for you etc etc, and I do get it that just because we don’t know the criterion just yet doesn’t mean we need to give up, still because the problem was identified, some changes to the original position must be made, don’t you think? Yet you go back to SR thesis being that there’s a world independent of human etc etc and it’s all back to the original set of issues.
John, are you suggesting that scientists are somehow subjectively involved in their scientific pursuits and that some scientific theories are a basic result of consensus rather than truth? I mean if it’s all up to specific individual scientists and their opinions, how can we guarantee that we will not have another Lysensko?
Here’s another example, closer to neuroscience: the construction of self-organizing neural networks in AI/cognitive science. The idea is to start with a flat array of “neurons” and a simple mechanism by which these neurons can construct “synapses” connecting each other. The task assigned to this neural array is to learn something about “the world” — some limited domain of knowledge specified by the researcher. The neural array scans data presented to it and generates tentative hypotheses about how the data clump together into patterns: this is a rock, that is a bird, that is a flock of birds, etc. The network then asks “the world” — i.e., the experimenter — if its tentative hypotheses are accurate. Based on repeated question-feedback cycles, the network organizes itself into a system that gets better and better at knowing things about the artificial world it confronts.
These self-organizing neural networks are built using statistical algorithms, including factor analysis, iterating across the nodes in the network. Depending on initial conditions of “the world” and the sequences in which the algorithms fire, a potentially limitless number of network structures can assemble themselves, each achieving comparable levels of success in knowing its world. Which one is correct; which one corresponds most closely to the real?
Of course we’re talking about simulations here. What happens when imaging technology advances to the level where it’s possible to track actual neural firing patterns propagating through actual brains? The question remains: what accounts for these specific patterns? The neural data are entered into a big data matrix; factor analyses and other statistical analyses are performed; alternative mathematical models are generated, many of which account adequately for the findings without being a perfect match.
Mikhail, we were writing at the same time. I’m saying there’s an iteration between observation of the world and interpretation of those observations. Some factor-analytic solutions really are better than others; i.e., they account for more of the variation in the data. So the weaker solutions are discarded and subsequent tests are invented to distinguish among the stronger solutions. Scientists are prepared to acknowledge their subjectivity in intervening in the ways the statistical algorithms process the data, in deciding whether a seemingly weak model might hold unrealized promise, and so on. But the complexity of the world always outstrips the complexity of the model explaining the world. So degrees of freedom always remain when the Fellowship gathers to decide which is the One Ring. At least it’s possible iteratively to eliminate the phonies, realizing that mistakes are sometimes made and good candidates are allowed to languish for generations at the bottom of a cave.
John, I think your model is more complex and closer to what I thought (from little that I know) science communities actually work like than, say, some other presentations of science in which scientists somehow simply and immediately knows stuff and we, philosophers, are idiots for all the arguments and all that…
John,
“These self-organizing neural networks are built using statistical algorithms, including factor analysis, iterating across the nodes in the network. Depending on initial conditions of “the world” and the sequences in which the algorithms fire, a potentially limitless number of network structures can assemble themselves, each achieving comparable levels of success in knowing its world. Which one is correct; which one corresponds most closely to the real?”
As I understand cog-sci, in order for these neural networks to be shown to correspond to the real, they have to be shown to correspond to some actual neurological structure in the brain. So the neural networks created can be fantastic for AI, as models of cognitive efficiency, but for cog-sci they have to match up to the actual ways in which we think. (e.g. the speed or slowness of our ability to do cognitive tasks, or the breakdown of the network into functional units which then map onto specific brain areas, etc.) So while I think you’re right about factor analysis in psychology, I don’t think the neural network example is as persuasive.
That being said, I still think your point holds – and as discussions of symmetry and beauty in physics show, it’s not clear that there’s not some rather large anthropomorphic biases seeping into even fundamental physics. It’d be interesting to hear how statistical analysis plays into physics as well – I recall reading the other day about how the LHC produces terabytes of raw data per second (see this article). Obviously that raw data needs to be pre-analyzed at some point prior to any human intervention.
“for cog-sci they have to match up to the actual ways in which we think.”
Agreed. The same could be said of mathematical string theory models, which precede empirical evaluation of these models vis-a-vis actual universes.
In cog psych, processing speed/accuracy, brain maps, actual neural firing patterns, etc. generate data sets that must be analyzed separately and correlated with each other. The analytical tools are probabilistic and will generate models that look a lot like the hypothetical AI/cog sci models. Results will almost certainly not converge on the single solution. But as data get more complete some hypothetical models will be eliminated from consideration, others will jump over the hurdle, new ones will get proposed. Iterative process.
“they have to match up to the actual ways in which we think”
The question at hand concerns what it means to “match up.” Surely task performance speeds and fMRI maps have only tenuous relationships with how we really think. They’re descriptors or indicators pointing to thinking, not direct representations of thinking. Multiplying the number of descriptors/indicators isn’t going to turn them into representations of the real. But I suspect you’d already agree with that point, Nick.
It seems to me that you are suggesting, in very sophisticated and nuanced ways, a very simple idea – it takes a human interpreter or a community of human interpreters to make sense of scientific data – fMRI images don’t really speak for themselves, they need interpretation – how is this any different from Galileo discovering the moons of Saturn with his primitive telescope?
Mikhail,
Return to the argument from the arche-fossil. What are the properties that allow us to infer the age of these fossils? The radioactive decay of carbon isotopes. This would be an example of a primary quality. As I see it, there is no way one can both support the findings of neurology, evolution, or geological time and be a correlationist. You simply can’t simultaneously hold that we only know phenomena (without ever knowing things as they are in themselves) that are constructions of our own minds and that our minds are products of nature.
Whatever those properties are, they are properties of an object, a fossil (even if we call it “arche-fossil” or “fossil-matter”) – you are looking at an object. Yes, I’ve heard the argument that since this particular object points out to a time before humans, then someone it is disproving correlationism, but leave it aside for now – you are holding an object that is no different that my cup of tea, it’s much much older etc etc but appears to me as having weight, shape, smell, taste whatever.
So you measured the age of this particular fossil and found out it is very old, in fact, it comes from “time before time” – it’s really old in “real time” – return to my previous objections vis-a-vis this nonsensical notion of “time before time”…
Primary quality of what? Carbon isotopes? Of the fossil I am holding in my hand? How do you know that this – radioactive decay – is a primary quality? What is the difference between measuring a length of a cup and declaring that it is a primary quality and measuring the radioactive decay of carbon isotopes and declaring it a primary quality? Both are physical characteristics of objects, are they not? And you previously stated that length is not a primary quality – what about the weight of your fossil? Is it a primary quality? Or is a primary quality something only a science can see and tell us about?
You can see that one cannot support science and be a correlationist all you want, that’s not the issue here, clearly we are disagreeing on a number of issues yet when the push comes to shove, you hide behind science and all its awesomeness citing various scientific curiosities. If we are going to disagree, we need to argue our points, as we were for some time now, without hiding behind supposedly self-explanatory scientific data. As much as you are fascinated with science, you are not a scientist and I personally can barely add and multiply without a calculator – can we discuss these issues as non-scientists? can you present the problem in a philosophical form that I and others who are not scientists can understand?
John,
By ‘match up’ I was thinking more in line with the studies of brain damage – a certain area of the brain is injured, and subsequent studies reveal that, say, working memory is hindered. So in that way, we uncover the various functional components that the brain is composed of (and where they are located), and then we build neural networks that attempt to model that system. All that being said, I like Churchland’s take on the issue of eliminativism – the reduction of mind to matter would be akin to the reduction of caloric substance to a kinetic theory of heat. People can always state that the kinetic theory of heat is really explaining how caloric substance functions (just like people can explain that neuroscience is really explaining how some immaterial thinking functions). But at a certain point, it’s realized that caloric or thinking just doesn’t exist in the way we thought it did – the evidence and power of the alternative theories just overwhelm the ‘folk’ theories. There’s no need to postulate something above and beyond. So while I don’t disagree with “multiplying the number of descriptors/indicators isn’t going to turn them into representations of the real”, it might turn out that fMRIs and other measurements, by measuring the brain, are actually measuring thinking in a very real (albeit low-resolution) way.
Mikhail,
Not to sidetrack the debate onto an old topic, but I admit I never understood your position on the ‘nonsensical’ notion of a time before time. Your claim that it makes no sense seems, essentially, to rely upon taking the Kantian form of time as the only meaning of time. Why could there not be a form of time produced by the brain, or imposed by a transcendental subject, and a physical, or realist, notion of time? Of course the latter would be different from our everyday sense of time, but surely that alone doesn’t make it nonsensical. (And I have no particular theory of realist time in mind, I just don’t think that the possibility is intrinsically absurd. Physics may end up showing there is no such thing as time, and even someone like Deleuze ultimately argues that a pure, static form of time underlies the sense of time moving. So it’s not clear to me that time only exists in the way we experience it.)
Nick, in our discussions with Levi, I objected to Meillassoux’s argument from “real time” or “time before time” – I did not rely on Kantian view of time at that point, I simply argued that Meillassoux appeared to be somewhat confusing in his discussion of time, as if time was something obvious and in no need of explanation. In other words, he already assumed that “real time” (and not correlationist notion of time) was basically the only correct version of time which made his criticism of correlationism very easy – the desk was stacked, so to speak. I don’t know if this short version of my objection makes any sense.
Look at his response to an objection on pages 20-22. While he spent a lot of time carefully constructing a rather strange vision of what correlationist would “believe” (misrepresenting Kant, for example, on page 15 with all that talk of givenness and such), all of a sudden you have notions like “time of science” that are never quite explained and in discussions with Levi I questioned such assumption, and brought up, for example, Leibniz-Clarke correspondence where the notion of “absolute time” (their version of “time of science”) is criticized and so on. Meillassoux, I think, conceives of absolute time as that which is already there, it is in this absolute time (“time of science”) that humans appear and their “conscious time emerges in time” – my beef with all of that was the way in which a rather strange notion of “time before time” comes into view but is never directly addressed. The whole point of an objection is based on it, yet it escapes the attention, as far as I can tell.
Okay, that makes more sense to me. I agree that the notion of absolute time that Meillassoux refers to is very underdeveloped so far, but that’s not to say that it’s intrinsically nonsensical or anything. It just needs to be fleshed out more (one would hope in his next work!)
I still fail to see how the decision that one set of presuppositions is binding can be reasonably justified by assuming what you set out to prove, that they are binding–that it is an “empirical question” whether everything is an empirical question. There is a moment of decision that can be justified ex post facto by appeal to eliminitivist principles, but not in advance, this is a moment of freedom not because of any supposed free will, but because of the fact that it exceeds any given system. We may not make the decision spontaneously, but it is still a decision–the law itself is not bound by law.
If science tells us that our brains produce illusions that determine the way we think, this is a self-refuting hypothesis because it undermines any chance of proving itself. If we can only think the way we evolved to, what makes us think we can think about the way things really are at all—including thinking about the way we evolved? Again, there is circularity–we are explaining something on the basis of the thing to be explained, assuming that science tells us how the world us and justifying this on the basis of science. The rightness or wrongess of eliminativism cannot be an empirical question.
My point is that his critique of correlationism in his argument of the “arche-fossil” depends on his “time before time,” therefore it cannot wait until the next installment – that is, the issue of absolute time is something rather essential if you want to argue that your opponent cannot account for something like ancestral statement while you are sneakingly bring in a completely unjustified (for the purposes of this philosophical argument) notion of “absolute time” knowing that so-called correlationist would operate with a different conception of what time is.
It would be like saying something like: “Let’s assume there’s absolute time, in this case, you cannot tell me how this or that is possible. therefore you lose!” – clearly, if a correlationist is really allowed to respond, the “let’s assume” part is not going to do it, therefore there is no refutation of correlationism that, for example, Levi heavily relies upon in his arguments. That was my main problem.
Mikhail,
I never understood your point about time before and now have a better sense of where you’re coming from. When Meillassoux uses the term “time before time” this remark only makes sense within the framework of Kant and phenomenology. What he is referring to is a time prior to the constitutional activity of consciousness. Within a Kantian framework we cannot make sense of the idea of a time belonging to things in themselves, but this is precisely what is required by knowledge statements about times that precede the existence of humans or life.
The contradiction we encounter between science and neurology, geological time, and evolution is as follows: Something that is a phenomenon from within a correlationist framework– the brain, geological times preceding the human, times prior to the emergence of life, etc –is functioning as a condition for mind. Yet mind is to function as a condition for all phenomena. You can’t have it both ways. This generates a contradiction at the heart of correlationist thought where we are given the stark decision of choosing between these sciences (and the massive bodies of evidence that support them) or the correlationist framework. I am not “hiding behind” or “retreating to” science, but merely pointing out that correlationism fails to do the very thing it’s supposed to do: ground knowledge.
Nor am I “assuming” an absolute time. Rather all the evidence points to the existence of a time prior to human consciousness. Who ya gonna believe? Kant or your lying eyes?
And yes, I am committed to the thesis that primary qualities can only be discovered through inquiry and that the philosopher cannot name them or know them a priori from our armchairs. This is part of the whole anti-correlationist position: that the world as it appears to us in our day to day lives shares very little resemblance to the world as it is in itself.
Mikhail,
It’s also worth pointing out that this is a logical fallacy:
You are suggesting that position x is true because your opponent doesn’t have a solution to a particular issue. This would be equivalent, for example, to suggesting that it is wrong to point out that a particular economic policy has all sorts of problems because that person does not have an alternative to these economic policies. The lack of an alternative doesn’t entail the correctness of the opposing position, only that more work needs to be done.
I also fail to see how or why Meillasoux can say that the arche-fossil preexists the correlation, and maintain that the arche-fossil is a scientific datum, when he jettisons the principle of reason which means that God could have planted the arche-fossil, it could have come into being ten minutes ago, or whatever.
What is the “principle of reason” that Meillassoux is “jettisoning”?
Actually, as I mentioned several times, this is not a “Kantian framework” – Leibniz already had issues with Newtonian absolute time (and so did, if I understand it, Einstein, but don’t quote me here) – let me ask you this simple non-scientifically phrased question: was there time before the Big Bang?
If these are my only choice, I’m going to believe Kant because my eyes are indeed very lying, things look very solid to me when I know that physically speaking a block of granite is as empty as a block of air – you in fact (together with QM) are sneaking in a concept of time, absolute time (as opposed to relative time) without ever explaining where it comes from, it just is, right? because “all evidence – self-explanatory evidence – just points to it” – c’mmon, I’m sure you can do better than that!
No offense, and I know you’re sensitive about religion and all, but this sounds to me like a confession of faith rather than a philosophical position – all kinds of objections can be and are continuously raised against this position, yet you keep reciting it as if it is a credo of your anti-correlationist faith…
No, I am not – read my objection again, I suggest that it is unfair to assume something that an opponent does not agree with – a particular notion of time – and then, based on this unshared assumption to purport to destroy the opponent’s position – where exactly is my fallacy?
Not really, it would be equivalent of accusing the opponent of being inconsistent in his position because he fails to account for an economic scenario which is formulated in terms he doesn’t find acceptable to begin with.
Let’s say you were trying to persuade me that torture is okay – hot topic and you started by saying: “Assuming that torture allows us to acquire valuable information…” – an assumption that I would contest – and then proceeded with your argument concluding that I somehow am unable to explain A or B when the whole premise, if it were to be stated that way, would have stopped the whole conversation in the very beginning.
QM does not do it that obviously, but he does introduce the idea of “time of science” without ever considering the implications, without really raising the whole issue of what time is – or maybe I’m missing something?
I like this line of argument, because it seems that it is precisely how you are trying to refute correlationism – just because correlationism cannot account for X, as you claim, does not mean that Y is true (in your case, some form of realism, naturalism, materialism with a dash of speculation)…
Levi says this:
Notice though, that a Kantian doesn’t actually say that the pure form of time is a condition of mind. A Kantian says that the pure intuition of time is a condition of the possibility of experience. Somehow we’ve managed to assimilate ‘conditions of possible experience’ to the ‘conditions of mind’ and — worse — ‘conditions of mind’ have been assimilated to ‘grounds of cognition’. There’s a pretty big difference among these notions here. It’s also worth pointing out that there’s a set of very good reasons for distinguishing among conditions for something, grounds for something , and determinations of something. The present discussion seems to level out these differences (in much the same was as Fichte does, actually). But who really wants to be a Fichtean? (Schelling on a bad day, is my answer). In fact, I’m plenty happy with an argument that makes Fichteanism less palatable. It makes the post-metaphysical Hegelians and Kantians look way ore respectable.
It’s also worth pointing out that “Geological Time” isn’t really different from “cosmological time,” or the two very different from “evolutionary time,” or “human time.” It’s all the same “time” — It’s just the interval or period (the metric, which is ultimately a human imposition, a way we make various kinds of shifts coherent to ourselves) that shifts in relation to the “rate of measurable change” that a particular discipline requires. Quite simply put, it’s a matter of perspective, not a question of qualitiatively different ‘times’.
Frank,
I may be misunderstanding your position, but let me say two things I think might help in understanding how the eliminativist position can avoid being self-refuting. One, if – hypothetically – we come to the point technologically where we can produce any type of experience we want in another person (e.g. by stimulating specific patterns of neurons), this is powerful (and empirical) evidence for the eliminativist position. But of course, while we can see this occur in someone else, we’ll never be able to experience our own brains as producing our own thoughts. Which leads me to my second point – namely, Metzinger’s work which shows how we can understand both the production of our everyday experience, as well as the impossibility of ever experiencing ‘being no one’, i.e. being able to personally experience our brain constructing ourselves (this due in part to what he calls the transparency constraint). So while eliminativism goes counter to not only commonsense, but also to our very cognitive constitution, the power of my first point is that it appears to be true despite that. You’re absolutely right though, that the issue involves a lot of circularities and is certainly tough to make sense of.
Mikhail,
I’d be careful (or maybe you really do mean it), but it seems like you are saying that cosmology, evolutionary theory and geological history are all essentially fictions.
Regardless of the specific details of all these theories, they do all point to the existence of a time before correlationist time (whether it be Kantian, Husserlian, or anything else). And really, I think that’s all Meillassoux’s argument needs to get off the ground. If we want to take these sciences seriously, then we have to admit that there is an alternative time to correlationist time. (We might call it a realist time, just to avoid any Newtonian connotations with absolute time.) In other words we have to think the time within which correlationist time emerges.
Nick,
My response was to Levi’s proposal that I choose between Kant and my own eyes – admittedly a half-joking one – which I take to mean something like: choose between philosophical arguments (say about time not being a substance, but a relation) or crude empirical data that apparently speaks for itself. I go for philosophical arguments.
I don’t think they point out the existence of time, I do think that they all assume a kind of absolute time, which is not the same thing – we’re not here discussing these sciences, I’m sure they have their arguments, we are talking about a lack of philosophical explanation on the part of QM, an explanation that, because his refutation of correlationism rests on it, one is not unreasonable to expect.
Again, the burden of proof is not on so-called correlationists – Kant has a perfectly fine (you don’t have to like it or agree with it) explanation of how it is possible both to think that time is a form of intuition and that physics actually empirically deals with objects and so on (empirical realism and transcendental idealism) – it is on QM and Levi, yet all I hear is stuff like: How do you account for arche-fossil taking that what science tell us about it indisputable? Ah you cannot, then you are damned to correlationist hell!
What I don’t get is why we need to take arche-fossil to make this argument? Why not take a simple issue of causality – for Kant, causality is also not something that is independent of mind and found in the interacting objects. Why not simply say – physics claims that when a billiard ball hits another billiard ball in such and such conditions, X is going to happen – how do you account for that, Mr. Kant? Posed this way, i.e. ascribing causal relation to the objects themselves, assumption Kant would not share, and then asking him to accoutn for X should be a good enough refutation of correlationsim, right?
Mikhail,
Taking you at your word here, I think this is actually pretty revealing – first, your preference for philosophical arguments is a major differend between us. I am more than willing to accept empirical findings into these debates. (And it’s good to realize where we just plain disagree.) And two, it sort of justifies Levi constantly bringing up science, as a counter to your preference for philosophical arguments.
I think, in order for those disciplines to make any sense whatsoever, they have to posit the existence of a time outside of correlationist time. If you want to argue that this absolute time doesn’t exist, then it seems to me that you are arguing against the very validity of any of these sciences. It may be an assumption of them, but it’s a foundational assumption. And the evidence from these sciences strongly points to a time prior to correlationist time (independently of any specific theory, e.g. the inflationary theory of the Big Bang). That may be an assumption of Meillassoux as well, but it’s an assumption I’m more than willing to subscribe to.
And this is why I think the arche-fossil is a more significant idea than causality – because the arche-fossil points to a time prior to the very existence of any sort of correlation. It indexes a time prior to any sort of empirical/transcendental doublet. Causality doesn’t do that alone; it doesn’t suspend the absolute authority of correlationism, making it only a contingent moment in a different temporality.
Fair enough, of course, I don’t think it is that revealing, I mean I always approached these issues from philosophical argument point of view – my point is that as a philosophically inclined person, I tend to think that empirical data does not speak for itself and needs an interpreter, or a community of interpreters, therefore I am skeptical when you say that you are willing to accept empirical findings as though they are somehow pure and interpretation-free ready-to-use chunks of data – but we differ here and I think it’s another conversation altogether…
My problem with that, as I stated many times, is that Levi is a philosopher, not a scientist, which is to say not that he doesn’t have appropriate expertise or anything of that sort, but that he approaches issues from a philosophical perspective of realism etc etc. Yet he throws around scientific data as if it somehow makes the argument in itself – data or examples are still in need of interpretation when it comes to their use in philosophical arguments – I’m not a scientist, I don’t know shit about specific scientific data or methods, but if you translate it into a philosophical argument, I’m all about it. That was my point. As I also commented to Levi, he seems to be seeking solace in Science (intentional capitalization), which is fine with me, but I constantly come off as some obscurantist anti-science weirdo because I dare to question something that is apparently self-evident such as “absolute time” which it is not at all.
Nick,
Thanks, and I think I see where you’re coming from, but it is one thing to say that empirical evidence can influence you’re accepting eliminativism, and another to say that it’s up to empirical science to show whether eliminitavism is true or not, which is incoherent largely because of reasons I’ve already stated. But if you want to say, as you have, that if science can show x,y,or z you personally will be convinced of the elimitavist position, I have no argument (at least at the moment) against that. If you’re not willing to restrict yourself to this, however, we don’t necessarily have to argue about it any more if you don’t want to (if you want to, I’m game, though).
Levi,
On page 71 of After Finitude, M says “..the principle of unreason teaches us that it is because the principle of reason is absolutely false that the principle of non-contradiction is true.”
I assume that’s what you were asking, not what the principle of reason is, but here’s a quote from Kant anyway just because it’s pithy and good:
“The principle of reason is…a…rule- prescribing a regress in the series of conditions for given phenomena, and prohibiting any pause or rest on an absolutely unconditioned.”
Not necessary, as Alexei pointed out above, all these “times” are different quantitatively, not qualitatively, if they are to be called “time” and they can be easily explained from say a Kantian perspective which I think Kant in fact did in the first Critique. I think what happens is not an assumption of time outside of correlation, but a simple lack of any distinction between “absolute time” and “relative time” – a kind of naive time of Newtonian physics – again, philosophical argument here clearly has a long tradition, and what the implications are for the sciences, I don’t know, I’m not a scientist…
Not necessarily – as Leibniz and Kant showed, the idea of absolute time and absolute space are contradictory – I mean if you take just Kant’s discussion of time/space, there’s plenty of things there. Now one can say: So what? Kant shows that time/space cannot have absolute reality, but clearly I see spacial dimensions, I can measure distance, I can count time and so on empirically, I must believe my lying eyes and disregard all that philosophical nonsense – in this case, how am I suppose to argue my point?
I don’t really know why we’re talking about Kant at this point – take Zeno’s paradoxes about space – is our universe spatially finite or infinite?
How exactly does it point out a “time” prior to correlation, if correlationists argue that “time” is necessarily a part of the very correlation – does it mean that there was no time before humans came around and started “timing” reality? Yes, if time is understood as a form of intuition. But that’s absurd, cries out (metaphorically speaking) QM, we all know what time is and it’s not a form of intuition, it is what science tells us it is or what science assumes it is when it does its thing!
Here’s the crux of the issue – QM assumes that we all agree that “time of science” is that commonsensical time we all know, the only one that just is, Newton’s absolute time/space which contains all the objects, so when he makes a correlationist say something like what I’ve said: “There was no time before humans arrived” he actually makes a correlationist says (literally, in his rejoinder) something like “There was no absolute time of science before humans arrived” and everyone laughs – but clearly you can see a quick substitution?
Sorry about multiple comments, your comment window is so small, I lose track if I write more than a couple or paragraphs
Of course he could be wrong, but here’s what Stephen Hawking says about his own scientific praxis:
“I would say that I’m a realist in the sense that I think there is a universe out there waiting to be investigated and understood. I regard the solipsist position that everything is the creation of our imaginations as a waste of time. No one acts on that basis. But we cannot distinguish what is real about the universe without a theory. I therefore take the view, which has been described as simple-minded or naive, that a theory of physics is just a mathematical model that we use to describe the results of observations. A theory is a good theory if it is an elegant model, if it describes a wide class of observations, and if it predicts the results of new observations. Beyond that, it makes no sense to ask if it corresponds to reality, because we do not know what reality is independent of a theory. This view of scientific theories may make me an instrumentalist or a positivist — as I have said above, I have been called both… In my opinion, the unspoken belief in a model-independent reality is the underlying reason for the difficulties philosophers of science have with quantum mechanics and the uncertainty principle……”
John,
So what would Hawking say about the arche-fossil and the “time of science”?
[...] Metaphysics, Ontology, Real, Speculative Realism No Comments In response to a terrific post by Nick over at Speculative Heresy, the debate surrounding correlationism has continued to swirl. I [...]
Mikhail,
From Hawking’s quote we might infer something like this… I regard the solipsistic position that nothing existed before humans as a waste of time. No one acts on that basis. But we cannot distinguish what is real about pre-human time without a theory. Obviously no one has direct experience of prehuman time. Empirically-supported theories concerning the carbon-14 decay profile and the speed of light enable scientists to use phenomena available here and now as tools for extrapolating the ages of prehuman situations and events.
Clearly, Hawking is some minor obscurantist and needs a good philosophical spanking, or worse, he is a closed correlationist because he apparently thinks that science cannot do without a “theory” – last time I checked, these theories come from human agents who are interpreting (good grief) and agreeing/disagreeing on certain matters – this is a step too close to claiming that there’s a correlation between observations and theory than I can take.
Mikhail,
I think we may be talking past each other to some degree, but let me try to clarify what I’m saying.
To be clear, ‘absolute time’ is not referring to Newtonian time. Einstein empirically discredited that (and Kant and Leibniz, as you note, philosophically discredited it). Absolute time, as Meillassoux uses it, is just a short hand for a time outside of the correlationist time (again, I’ll take Kant and Husserl as being the archetypes of this view).
Now when I say that absolute time is a fundamental assumption of cosmology, evolution, etc., I mean that these sciences are speaking of a time before the very possibility of correlationist time. To deny that an absolute time outside of correlationist time exists, is to deny that these sciences are speaking about anything. They literally make no sense if we assume time (and really, existence) burst onto the scene with the emergence of thought. But to argue that absolute time exists is only to accept a very minimal definition of it – that correlationism emerged within something larger. What that something is, is undetermined so far and a problem for future work. But that it is, seems indisputable to me. (And I believe Hawking’s quote says no more and no less than that, as well.) But maybe this is another manifestation of our differend, since I take these empirical sciences to clearly show the existence of an absolute time, whereas you are more focused on the philosophical conundrums?
The problem for correlationism then, as Levi succinctly points out in his post, is that correlationism sees the mind as condition for Nature, whereas the existence of absolute time shows Nature to be the condition for mind. (Although I’d need to read Kant’s later work to see how the Opus Postumum fits into this schema. I do have Forster’s book, on your recommendation, which I should really crack open.)
As for Hawking’s quote, I think he’d need to respond to the idea of structural realism. No one is denying that theories are used to give us knowledge about reality. What the instrumentalist says is that these theories are only pragmatic and have no truth-value, whereas the structural realist will say that this is incapable of explaining the predictive success of science.
Nick, I think I understand your position but the problem with Meillassoux’s argument succinctly is this: while it looks as though he is critiquing correlationism from “inside” by showing how it cannot account for something like arche-fossil, he in fact is critiquing correlationism from outside perspective by imposing the meaning of time on correlationism that it would not accept. As I tried to show, his refutation of correlationism rests on the assumption that correlationism does not share, i.e. that time is something that is a property of mind-independent world.
Now I think you are buying Meillassoux’s argument about “time before time” too easily. There’s no two types of time – “absolute time” and “time of correlation” – it is either one or the other, that’s the point. If you think time is something that is characteristic of things-in-themselves and human subjects only recognize this time and learn to measure it, then there’s no “correlationist time” – if you think that time is a relation that the mind establishes (to put it simply) between objects or events (this comes before that etc etc) the same way space is a relation, then there is no other kind of time, period. I don’t think it disproves anything about cosmology or geology, as Alexei mentioned already, that time can very well be understood in terms of “correlationist time” – Now the question whether time existed before human givenness, as Meillassoux puts it, only makes sense if you already rejected “correlationist time” – that is before you are even making your argument against correlationism, you’ve already discarded its major premise. That’s all I was trying to show.
Since Levi does not want to address this issue, let’s have a thought-experiment (without any scientific engagemen yet) – did time exist before the Big Bang? If you say “yes” then the question is not only “how can you know it” but also “have can you think that time before absolute scientific time?” Is time infinite then? If you say “no” then the question is “how can you think that lack of “time before time” – are you imagining an empty black static what? These are legitimate philosophical questions and I am very confused by Levi’s insistence that we sit back and let science answer all these questions – I find it not only absurd, but disturbing, coming from a philosopher.
Bottom line, I’m trying to show that Meillassoux’s argument is not as neat and tidy as he thinks it is and I take Hawking’s quote to say that because the so-called commonsensical notions (like mathematical time) are so widely regarded as “self-explanatory” people have hard time thinking about quantum mechanics – we are used to think of time as a property of objects, as a sort of container in which all objects are and interact, but one might very well argue that such an idea of time is specific to modern post-Newtonian science…
Correlationism, as far as I can tell from Meillassoux, is insistence on correlation of thought and being and not on dominance of being by thought or thought by being. In this sense, the above formulations – “mind conditions nature” and “nature conditions mind” – are both correlationist formulations, because they insist on the interdependence of two terms, i.e. their correlation.
Let’s say for the sake of the argument that “mind conditions nature” is a correlationist position and is false, then “nature conditions mind” is outside of our ability to demonstrate to be either true or false, because a passive mind cannot actively discover that it is being conditioned by nature. Kant’s position (since he is a representative here) is precisely that these two formulations are giving us either rationalism/idealism or empiricism, and he rejects both in his project – this is why his philosophy of nature is so interesting because it is ultimately a test ground for his whole philosophical system, and that is why he needs things-in-themselves.
I posted a link to one of Forster’s essays on Opus Postumum at PE (I don’t know if your mention of Forster is in relation to that post, but assuming it might not be) in which he gives a pretty good summary of Kantian project of philosophy of physics.
Just a couple of things, which may be totally insignificant.
RE “the time of science.” As Far as I know ‘time’ isn’t an ontological entity for the sciences either (I tried to make this point above, but clearly didn’t succeed). Time = metric/mathematical dimension in virtue of which rates of change/movement can be calculated. Although theoretically ‘primitive’ (i.e. unanalysable), it’s not an object, or structure. Time isn’t a property of anything — nor is space for that matter (See the notions of block time and Spacetime for instance — more here). I take it that this is Mikhail’s point, really. If you don’t have theoretical objects that can undergo certain kinds of transformations, you don’t have time (hence why the big bang has come up again). So honestly, much of this argument seems to be based on a distortion of how the sciences actually handle time. It’s never treated as a property, but as a dimension (which means, it’s not something objects have). Ultimately, what’s important is that time and space are a priori — independent of experience, necessary, and non-conceptual — so they are by definition independent of facticity.
Less philosophically put, the Math comes before the observation, and actually determines what it is we’re looking for, what can be observed, and what follows form it. why isn’t that enough? I mean, this is actually one of those moments where we should actually look at what scientific researchers do with time, rather than speculating about what time is independent of us humans.
I would like to officially reveal that Alexei and I are actually the same person – I don’t see how else I can explain our almost nauseating agreement about everything, except Alexei’s absolutely ridiculous unjustified creationist anti-scientific reactionary notion that somehow Bruins are going to make it into Stanley Cup Finals this year – what a preposterous and perverse notion!
dammit Mikhail — you’ve blown our/my cover! You’ve ruined a great grift.
But seriously, the Bruins are going to the finals. So quit trolling me….
[...] be conducted. Responding to Mikhail, Nick gives a succinct summary of Meillassoux’s argument, writing: I think we may be talking past each other to some degree, but let me try to clarify what I’m [...]
Clearly a Bruins fan has to believe the true nature of reality is a matter of faith and not knowledge. I don’t think there’s any other explanation for such an absurd position that defies all empirical knowledge.
Alexei,
I’ve been hesitant to talk about blocktime because (a) it’s highly contentious, (b) I don’t understand it well enough to even try to defend it, and (c) it pretty radically disrupts even the minimal terms of debate that we’ve set out (e.g. how would correlationism or Kant fit into a blocktime universe? Doesn’t it entail knowledge of the absolute?). I think it’s absolutely a worthwhile discussion to have (and not insignificant in the least), but an in-depth discussion would have to wait (at least on my end).
Mikhail,
I can see your point here, and I have a better sense of where you’re coming from now. But two things: one, I think different assumptions are the ground from which different philosophies emerge. Hegel and Kant have different assumptions, and each of them can easily critique the other for not sharing them. Which is not to say that any particular assumption is necessarily good, only that the claim that different assumptions aren’t shared isn’t sufficient to make an assumption bad.
And two, I don’t think the arche-fossil is the sole argument Meillassoux uses against correlationism. The other is the more properly internal one, where he roughly argues that strong correlationism, in order to refute absolute correlationism, has to uphold the absolute facticity of the correlation, thereby revealing the necessary contingency of the absolute. (I take these positions from QM’s discussion – page 35-42, and then 54-59.)
I’m willing to accept this – that there’s not two types of time – but I think even less contentiously, it can just be left an open question. To clarify my position better, these two types of times are more akin to two concepts of time, rather than two ontologies of time. The reference to correlationist time is then just a reference to the concept of such a time, rather than entailing that it exists alongside absolute time. It may turn out that absolute time encompasses and explains correlationist time; or it may not, and we’ll be left with a temporality that passes in a single direction, and some other type of temporality. (For example, blocktime would be like this – physically no such thing as time exists, but phenomenally it still occurs.)
The answer to your Big Bang question would of course be that we can’t know (yet, at least). So it’s a meaningless question at the moment. But how this differs from the correlationist time question is that we have evidence and reasons for believing in something prior to the existence of correlationist time. I think you may be putting too much substance on to QM’s notion of absolute time. If, as Alexei notes, scientific time is measured by rates of change, then QM’s hyperchaos entails that his notion of absolute time is distinct from scientific time (since hyperchaos would make it impossible to measure anything). So QM’s notion of absolute time isn’t correlationist, nor is it scientific. As of right now, it only makes a claim that there is something prior to the correlation. (Blocktime would of course refute even this, but it’s also problematic for correlationism.)
As for the Kant stuff, let me go and read Forster’s book, and I’ll see how Kant approaches the question of Nature conditioning mind. (Which, per your critique, might better be phrased as Nature independently conditioning mind, to show it’s independence from the correlation.) I actually got the recommendation from you a long time ago (I think in a discussion of Schelling), but just haven’t had a chance to look into it much more. Which reminds me too – your recommendation for Heinrich’s book Between Kant and Hegel was fantastic; if you have any others like that, I’d definitely appreciate them!
“Now I think you are buying Meillassoux’s argument about “time before time” too easily. There’s no two types of time – “absolute time” and “time of correlation” – it is either one or the other, that’s the point. If you think time is something that is characteristic of things-in-themselves and human subjects only recognize this time and learn to measure it, then there’s no “correlationist time” – if you think that time is a relation that the mind establishes (to put it simply) between objects or events (this comes before that etc etc) the same way space is a relation, then there is no other kind of time, period. I don’t think it disproves anything about cosmology or geology, as Alexei mentioned already, that time can very well be understood in terms of “correlationist time” – Now the question whether time existed before human givenness, as Meillassoux puts it, only makes sense if you already rejected “correlationist time” – that is before you are even making your argument against correlationism, you’ve already discarded its major premise. That’s all I was trying to show.”
I think a major stumbling block in these conversations has been the confusion of Meillassoux’s extrinsic motivation for realism in the form of the arche-fossil with his intrinsic argument against correlationism. If Meillassoux’s “argument” against correlationism amounted to a mere assault of transcendental and phenomenological philosophy with the “crude empirical data” of science, it would be entirely sound to ignore it on the basis of its logical circularity.
However, it is not at all Meillassoux’s intention in the first chapter of AF to disprove correlationism. The arche-fossil, as I read the text, is a sort of maneuver of guilt. “You think we can only know things as they are for-us? Well… okay, but it looks like you’ll have to deny the literal meaning of scientific statements in order to remain consistent. Have it your way, correlationist, but golly!, turns out that you’re no better than a Bible Belt creationist!”
If you don’t care for scientific realism than there is nothing threatening in the “challenge” from the arche-fossil. QM is simply highlighting the irony that the very same transcendental philosophy which sought to ground science is now incapable of acknowledging its results. That’s all.
Meillassoux’s true argument for “absolute time” and the “principle of factuality” assumes the central premises of correlationism and then tries to show how those very same premises necessarily collapse into a non-correlationist or realist philosophical position. It goes something like this (I’m taking this primarily from pages 54-60 of AF):
1. Correlationism claims that we cannot know the in-itself because of our finitude as subjects.
2. However, in accepting our essential ignorance about the in-itself, we are forced to also accept that the in-itself is primarily unknowable because it could always be other than our conception of it.
3. Yet, our knowledge of the contingency of the in-itself cannot merely be for-us because it indexes possibilities separate of our own existence, e.g. our absolute dissolution in death.
As such, correlationism auto-destructs when its insistence on our inability to disqualify any hypothesis about the in-itself because of our finite nature as subjects allows us to discover in this ignorance the absolute capability of all things to be otherwise.
And then, of course, from that discovery of radical contingency, Meillassoux attempts to prove the law of non-contradiction and the fact that something must always exist, and seems to hope that he’ll be able to derive a number of mathematical theorems from it in future work.
That’s what I understand Meillassoux to be doing in AF. He demonstrates that science is apparently incompatible with correlationist theories of knowledge and then strings together an independent argument against correlationism based upon the supposedly explosive consequences of its own assumptions. QM claims to have discovered the absolute in this attempt, i.e. contingency, and then uses it to found knowledge of the in-itself.
Sam, I’m in complete agreement. Your comment is a much more thorough and developed argument about the second point I was trying to make to Mikhail above. The focus on absolute time is secondary (philosophically) to the internal argument you outline. Reading absolute time in terms of guilt seems plausible to me too (at least as a motivation for QM).
I’ll have to go back and reread the remainder of AF again, but I have to confess that I feel his argument goes down hill after the second chapter. While I have the greatest admiration for the ingeniousness and rigor of his argument, 1) I just don’t know that I am willing to go all the way to the conclusion that the universe is absolutely contingent in the sense that it could change at any moment, and 2) I feel that his argument establishing this contingency is “speculative” in the negative sense of the word, proceeding through a priori reasoning than anything grounded in investigation. It is truly refreshing to see such a tightly argued work in Continental philosophy, but I have a great deal of difficulty following him to his conclusions.
I do think that the first chapter of AF is much more than a “guilt trip”. The argument from the arche-fossil poses a significant challenge, I think, to the claims of correlationist thought, requiring some account of how it is possible for correlationism to deal with such claims without simply dismissing them as dogmatic.
I fail to see how a Heideggerian is forced to treat scientific knowledge as a fiction. Does anyone actually believe that, for Heidegger, the mind generates nature? Or that nature doesn’t condition mind in any way? Or what? It’s all so vague, “correlationism” per se doesn’t exist, I prefer to say “Heidegger.” And I don’t see how the arche-fossil can be accounted for in the terms of spec-mat when the principle of reason has been proclaimed false, thus the arche-fossil could have spontaneously emerged 30 seconds before I found it.
Sam, fair enough – in my comments I primarily talked about QM’s specific refutation of correlationism through a an arche-fossil test (chapter 1), since Levi was so sure that it was the only argument we needed, if I suggested somehow that my issues with his “time before time” and this particular refutation somehow mean that the whole book is worthless, then I am here to correct that impression as I by no means was trying to draw such conclusion. In fact, I’m glad that you moved this discussion further into the book as I think there’s clearly much more interesting stuff there than a silly (I still think) arche-fossil thing. Still, I feel as if my point about “absolute vs relative time” is not clear enough – once I recover from all the psychological abuse that Levi is unleashing at his blog, questioning my every motive and making me say horrible horrible things, I would like very much to return to your (and Nick’s) comment and respond to it.
Levi,
I don’t disagree with you from my own personal perspective, although I think QM would disagree with both of us. I think the arche-fossil is more than just a guilt trip as well (hence my qualification in the comment above), although within the limits of QM’s work, I don’t think it does all the philosophical work we might want it to.
Basically, what I see QM’s work doing is to show the incoherence, internally, of the correlationist position. That is to say, even if we try to be correlationists, we’re logically led to a form of speculative realism. And if correlationism is self-refuting, in that specific sense, then we have to turn to some form of realism. Which is where we all tend to go our own ways. It’s as though the island upon which we’ve all been standing (correlationism) has sunk, and now we’re all swimming in separate directions hoping to find solid ground again.
Frank,
I mentioned this to Mikhail elsewhere, but I won’t assume you’ve read it. Basically, ‘correlationism’ is a short-hand for a lot of more in-depth details. Levi and QM himself have both outlined a number of variations on the correlation (including Heidegger), so while it can seem like a vague term, most of us using it are presuming all this background knowledge as well.
Your question about how the arche-fossil, as a scientific finding, can be reconciled with hyperchaos is a really good one though. I personally don’t know how to answer it, and it’s the type of question that makes me think we ultimately can’t hold onto this idea of hyperchaos. Presumably, though, QM will attempt to answer that question in his next work, since he does explicitly say that he wants to account for how scientific statements about things like the arche-fossil are possible. (A very Kantian approach, if I may add.)
Frank,
I don’t know that anyone is claiming that correlationism treats scientific thought as a fiction. Certainly that’s not the case with Kant. Rather, the debate is whether or not scientific knowledge claims reveal properties of things-in-themselves or only things-for-humans. I think it’s extremely difficult to argue that the former is the case with Heidegger given his account of disclosure and Dasein with respect to being.
Mikhail,
I fail to see how I speculated about your motives in either of my posts. Rather, I laid out my understanding of the arguments and tried to respond to your criticisms. That’s what take place in a philosophical debate. Why does everything always get so pissy and personal with you in these discussions? Why do you get upset when someone fails to understand what you’re claiming or tries to respond to your claims? It’s really counterproductive and exasperating.
Clearly, Levi, it is all my fault, I am sorry – it seems that I always get into these personal pissy fights with everyone I argue with, must be my “thorn in the flesh”…
Nick, Henrich’s book is really one of my favorites – it’s really sad that his work is not very well represented in English, he is one of those German scholars with enormous intuition but also very un-German (no offense, German readers) clarity and wit.
If you liked Henrich, you might also enjoy (and I really mean enjoy) Frederick C. Beiser’s books like The Fate of Reason or The Romantic Imperative.
Let me clarify:
I don’t think that the arche-fossil in AF functions “merely” as a guilt trip for the correlationist either. It’s obviously quite a scandal for the intellectual relevance of correlationism if it cannot take proper account of ancestral statements without rendering them meaningless. However, Meillassoux himself does not think this is enough to refute correlationism. As clearly stated in Collapse III, Meillassoux thinks that the Fichtean argument against realism, i.e. no knowledge of x without a positing of x, is a sound one. Therefore, the only way out of correlationism for QM is by way of internal or immanent critique. Scientific results may lead us to non-correlationist philosophies but, and I’m only interpreting QM here, cannot themselves justify a realist position.
Levi and Nick (and Brassier, Laruelle, Harman, and Grant, for that matter) probably don’t agree with this, but it is quite central to Meillassoux’s understanding of his own work.
Sam,
That makes more sense. This is what Graham takes Meillassoux to task on, but I Graham hasn’t struck me as forthcoming on an argument as to just how one is to get out the Fichtean circle. Laruelle then appears to be the only way out with his “non-philosophical critique” of philosophy, revealing the work of an arbitrary “decision” at the heart of every philosophy, but in my view, while I find his argument deeply compelling (what I understand of it any way… Nick and you’s, get to work and explain the rest to us in PLAIN language) this move doesn’t seem to get us very far as we remain in a sort of correlationist circle but now with the freedom to mix, move, and choose between different correlationist positions (Nick, please correct me if I’m completely butchering the “cash-value” of Laruelle here).
The other option, not very satisfying, would be something like a “paradigm shift” from a pragmatic point of view. Following Rorty’s strategy– as I recall it, it’s literally been two decades since I read any Rorty –here the strategy would be that we simply cease asking certain types of questions or doing philosophy in a particular sort of way. Take the transition from Scholastic thought to Modern thought. It’s clear that Modern thought fails to respond to a number of the problems of Scholastic thought or to resolve the problems that were at the heart of Scholastic thought. Moreover, it’s clear that Modern thought failed to address a number of criticisms directed at Modern thought in its own time. The triumph of Modern thought did not consist in refuting or proving it’s positions with respect to Scholastic thought, but simply in moving on and no longer asking those questions or doing in philosophy in that way.
If philosophy is approached instutitionally and sociologically, it seems that this way of transitioning between different orientations is fairly accurate as to what actually takes place. Those of us who went through grad school in philosophy departments will recall those tenured Emeriti who were still swept up with existentialism, the early phases of analytic philosophy, etc. They were looked at as curious fossils that you might study with to learn a bit about a particular historical moment in recent philosophy but who were no longer engaged in contemporary discussions. It seems to me, if my understanding of Laruelle is at all accurate, that this would be one of the consequences of his position… That philosophies are not so much “refuted” as simply discarded.
Nick,
I have read it, my point is that I don’t think “correlationism” really names a position and there’s a little bit of a shell game (not necessarily intentional or sinister, maybe I should say “shell-game-effect”) with names and positions out of which it’s cobbled together. The reason this is a problem is that it’s almost impossible, or perhaps is impossible, to DEFEND “correlationism,” so Mikhail goes with Kant and I’ll stick with Heidegger.
Levi,
I posted my last before I saw your last but one, which gets you out of the problem I pointed out with the arche-fossil, and Nick seems to want to opt out also. That’s all fine. As for the “fiction” thing–I got this out of one of Alexei’s posts, and now that I just tried to google it I realize I may have been duped, and it may be a paraphrase, not a quote–and perhaps an unfair paraphrase at that:
“If you’re a correlationist,” says QM, “you must think that the big bang is a fiction, or not be consistent!”
So, that one is my mistake, if QM didn’t actually say it. I’ll rephrase (and perhaps partially rethink) my objections (but they’re not based on the quote, I read the book, but a while ago now).
Frank,
I think it’s absolutely fair to say that correlationism doesn’t really name a position – I’ve intentionally gone to Kant to have some sort of actual philosophical position to rely on and I think it’s more than fair to do the same with Heidegger. At the same time, I think one can imagine a correlationist position the way QM wants to describe it, just for the sake of the argument, I think…
Frank,
No, I’ve come across no claim by QM to that effect. There are three options: To give a correlationist account of the big bang, evolution, etc., to reject correlationism as unable to give such an account and embrace realism, or to reject the claims of evolution and the big bang. QM does make some nasty comments to the effect of correlationism being a sort of sophisticated creationism because he doesn’t think correlationism can meet the requirement of the first option. But he provides arguments to support this point.
Frank, the line you quoted form me was indeed a paraphrase and not a quotation. For future reference, if I don’t supply I page reference, it’s not an actual quotation. But it is nevertheless accurate. See AF pp. 15-17. IN principle, I should quote that whole tract of text since it has to do with retrojection and the two levels of meaning, but a few choice excerpts should do. When speaking of a Kantian approach to ancestral statements, QM writes
And,
And,
Finally, from the same page as the last,
So all these quotations aside, I think a nice, synthetic reformulation is my paraphrase: If you’re a correlationist,” says QM, “you must think that the big bang is a fiction, or not be consistent!”
Alexei,
Aren’t you conveniently subtracting all the intervening steps in Meillassoux’s argument as to how he arrives at this conclusion, though? That is, aren’t you, as the defender of intelligent design might do, quoting all these things out of context so as to establish your characterization of his claim as “outlandish”? I’m more than happy to conclude, with Meillassoux, as a former correlationist– and there’s really nothing worse than a new convert, I confess –that the correlatioist position is equivalent to a creationist position.
Or to put it succinctly, the correlationist that denies this conclusion simply isn’t following the logic of their own position. Given the number of red herrings or smokescreens and the preponderance of ad hominems from on side on this discussion, I think he’s right here. I still haven’t seen a non-question begging address to this core issue and have seen a lot of subject changing.
Levi, I told you where to find the material. Make of it what you will. For my part, I’m confident in my interpretation.
[...] three cases, with an eye towards why the readings of Kant matter. (I won’t address the recent hot topic concerning time and ancestrality, since I can’t devote the energy to it, especially as [...]
[...] on! Anyway, back to the issue at hand. Over at Speculative Heresy, Alexei draws attention to a passage series of passages from Meillassoux’s After Finitude and concludes that he is basically [...]
[...] response here, plus a couple of exchanges at Larval Subjects and a continuation of the thread at Speculative Heresy. Things are heated but ultimately rewarding, I think, at least we’re [...]
What is a red herring? It sounds like some sort of fish and keeps popping up here and there, but I have to say, excuse my cultural ignorance, I find it odd…
Maybe the train has left this station, which gives me some time to regroup and consolidate my own thoughts. On p.3 of After Finitude Meillassoux states his twofold thesis:
“on the one hand, we acknowledge that the sensible only exists as a subject’s relation to the world; but on the other hand, we maintain that the mathematizable properties of the object are exempt from the constraint of such a relation, and that they are effectively in the object in the way in which I conceive them, whether I am in relation to the object or not.”
In reading your post, Nick, I understood you to be affirming the idea that the real consists of patterns, which can be discerned mathematically. You said:
“individual things don’t exist. Rather, what exists are ‘real patterns’ – temporal and spatial patterns which are mapped by the mathematical structure of scientific theories… (a) such patterns exist independently of any observer, (b) the notion of real patterns is determined in informational terms and is in principle available to computers capable of drawing natural laws out of raw data”
This idea corresponds closely to the second half of Meillassoux’s thesis. I intervened by observing that computers can extract mathematical patterns that do exist in the data without human sensible relation, just as Meillassoux asserts. These patterns are virtually limitless in number and may offer good statistical power as predictors of other phenomena. However, most of these patterns are rejected as spurious; i.e., they don’t seem to have any theoretical value in explaining or predicting anything. This doesn’t mean that none of the patterns is scientifically useful; it just means that many aren’t. Further, the many statistical patterns extracted from a single dataset can be arrayed from more useful to less useful rather than divided in binary fashion between useful and useless. This observation implies that mathematical patterns identified in the world might contain an admixture of the real and the unreal.
I skipped over another part of your post that addresses the first part of Meillassoux’s thesis. You referred to Ladyman and Ross’ interpretation of how information is received by observers:
“not in terms of direct perception, but rather as being ‘informationally connected to’ – meaning there exists an information channel from the X under observation to the observer P. Note again, that in formulating it in terms of information, Ladyman and Ross avoid any anthropomorphic bias in regards to observation. It need not be a visual observation, and it is in principle open to an infinite number of variations of information connections… in this formulation, any and all biases between the ostensible independent pattern and the observer are all incorporated into the informational network. So, for example, the theories which lead our attention to certain patterns are part and parcel of the informational connection; i.e. they are not a bias to be removed. In addition, as Ladyman and Ross state, “when we think about a relation of informational connectedness between some X and some P, we are thinking about both X and P as points (nodes) or regions (interconnected sets of nodes) in networks.” (308) In other words, this networked, informational form of epistemology negates any radical division between an observer and the observed. Or to put it in Laruellean terms, we are always already within the real. Both the observer and the observed are situated in an ontological framework underpinned by multi-scalar real patterns defined in non-anthropomorphic informational terms.”
I think this idea has merit in a couple of ways. Most directly, it presents a naturalistic account for breaking down the divide between observation and theory. Confronted with an array of mathematical patterns extracted from data, the scientist imposes her own pattern-matching heuristics to distinguish the relatively more scientifically useful patterns from the relatively useless. I’m reminded of Jackson Pollack’s process of “post-painterly selection,” whereby he divided paintings generated by random splatters into more versus less artistic results. Certainly scientists can articulate, at least in part, why one factor-analytic mathematical model is better scientifically than another. Some of it is personal taste, but there’s also considerable intersubjective convergence among scientists. These cognitive judgments regarding theoretical “goodness” emerge from patterns activated in scientists’ brains mapped onto patterns extracted from data sets. I’m down with that. On the other hand, there’s still the difficulty of asserting that the scientists’ mental patterns are capable of discerning the real from the unreal when evaluating mathematical patterns. Scientists work on themselves to purge their heuristics of subjective biases, but this work always remains incomplete.
The other important issue about data patterns has to do with the relationship between observable phenomena and the world outside of observation. Mathematical patterns reduce apparent phenomenal randomness to regularity. More importantly, patterns also consist of structured elements occupying a sort of intermediate scientific space between raw observable phenomena and pure theory. It’s presumed that the pattern is the source of individuation, the machine that produces the specific stream/array of phenomena that happened to pass in front of the observer during a particular interval. Set up another observation point and different phenomena would appear, but these phenomena too would have been pruduced by the same underlying pattern-machine. The scientist can even operate the pattern-machine intentionally to generate phenomena — this is how certain kinds of experiments are performed.
Scientists really do presume that “good” patterns aren’t just artifacts of mathematical simplification algorithms, but that the patterns actually produce the raw stuff of observation. This in effect is what Meillassoux asserts in his thesis. Empiricism isn’t “raw;” i.e., it’s not limited to the world of observable phenomena: I agree. However, Meillassoux takes the step that Ladyman and Ross also take: that the mathematical patterns identified by scientists are real. This I think is an unjustified leap. Patterns are “good” if they generate or predict observable phenomena (and if they’re “informationally connected to” scientists’ tacit cognitive heuristics for identifying goodness). But there’s still no assurance that a good phenomena-producing engine really works in the same way as the natural engine. A good pattern bridges the gap between observation and the unobserved, but the pattern functions scientifically as a useful simulation of natural processes rather than as a direct translation of the real thing. A means of specifying the strength of correlation between simulation and real remains to be formulated.
Mikhail,
Thanks for the references! I have The Fate of Reason, but I’ve only skipped ahead and read the ending so far.
Levi,
Just a brief comment on Laruelle. I’d resist saying that he leaves us still in the correlationist circle, with only the option to choose between philosophies. I think Laruelle’s basic position is that we’re already within the real (see his conversation with Derrida for an explicit statement to this effect). What Laruelle alone (distinct from non-philosophy) does is to show that any philosophy presumed to encompass everything is reliant on an unjustifiable decision. That by itself doesn’t do much, as you note. My position is that non-philosophy (where ‘non-’ means something similar to the ‘non-’ in non-Euclidean geometry) is wider than Laruelle’s particular system, and can eventually provide us with resources for thinking non-correlationally. (Laruelle himself admits this at one point, and I think it’s also the general basis of Brassier’s work.)
John,
I think I see better now what you were getting at. One of the aspects I left out of the discussion of real patterns is that Ladyman and Ross argue that real patterns must be projectible into the future – they must make accurate predictions of how the pattern will continue. So this disqualifies a lot of the patterns that can be extracted from a data set. I don’t know how they would understand the refutation of theories within psychology, though, considering it’s not a mature science in the same way as physics or chemistry. But the refutation of a particular pattern’s existence in physics appears to be recuperable within a realist framework because of the emphasis on structure. I don’t know whether a similar thing could be said of psychology though. (Ladymand and Ross only briefly mention psychology in a number of areas, preferring to focus on more mathematized social sciences like economics.)
I think this also somewhat responds to your point about determining the real from unreal patterns generated by factor analysis. Namely, in physics it seems quite plausible that structure is retained, revealing that we are definitely getting something right. The same might be able to be done for psychology, but I think it would entail more in-depth knowledge of the discipline than I have (but it’s something you might be well-positioned to do!)
I think this is accurate for Meillassoux, but Ladyman and Ross are careful to note that mathematics only describes real patterns. The patterns themselves aren’t mathematical (at least not in the same sense that being is mathematical for Meillassoux or Badiou). The theoretical and mathematical patterns we use to do science are themselves more like windows onto the real patterns – things that focus our attention onto real patterns at multiple scales.
Beiser is good, it reads like a novel (or maybe I’m just lame like that), you should give it a read, if you get a chance, maybe on vacation or something – I read Being and Time on the Black Sea beach back in the day, I didn’t like it much back then but still to this day when I pick it up, I think “beachness of the beach” (and shish kabob too, those Georgians know that stuff, that was back in the day when Georgia was part of the Soviet Union)…
“The patterns themselves aren’t mathematical”; the mathematical patterns are “things that focus our attention onto real patterns.” Yes, that’s a good description of how I see these things. I’m less persuaded by the window analogy. I suppose one could say that the incremental improvement of scientific models is an act of cleaning and polishing the glass so it distorts less — kind of Pauline or Heideggerian. A simulation can get better and better without ever becoming that which it simulates. However, incremental improvement in the simulation depends on having more precise knowledge of what the real thing does and how it does it. I suppose that’s what keeps me from abandoning the idea that our knowledge of the real is always colored by our knowledge of the real. But I’ll keep polishing my windows in hopes of seeing more clearly in future discussions.
“real patterns must be projectible into the future” — right, this is what I was grasping for in speaking of mathematical patterns as individuating machines. If the pattern merely describes the specific phenomena collected in a specific dataset, then the model is too closely tied to raw empirical observation. A mathematical simulation of real patterns ought to be able to generate/predict observable mathematical regularities in observations of new phenomena. Scientists presume that the pattern is generating the real, of which observable phenomena are merely a subset.
John,
Related to some of the stuff you were talking about here, you might find this lucid discussion of data analysis in particle physics interesting:
http://www.quantumdiaries.org/2009/05/05/data-analysis-in-simple-words/
Thanks Nick. I hadn’t realized that some of these same probabilistic techniques used in constructing personality theories and simulating neural networks are being applied also in subatomic physics, but it makes sense. “Reconstructing the objects in the event… to automatically produce (induce) models, such as rules and patterns, from data” — this sort of endeavor is important regardless of whether the uncertainty resides in the minds of the scientists or in the phenomena under investigation.
[...] started to work on. I think it will be helpful in settling some of the fascinating realism debates taking place in the philosophy blog world right now, especially in regard to Heidegger’s position on the [...]