Let’s have a new topic, preferably one that is not Christian apologetics.
This is mostly intended as a response to a comment by KN, but I think it deserves its own thread.
There’s a recent blog post elsewhere that is related:
- Knowledge and Reality (by Dan Kaufman)
Personally, I think of myself as a realist. But I agree with some of Dan Kaufman’s criticisms of traditional views of reality.
Now my response to KN. The quotes will all be from KN’s comment (linked above).
I think I’m more inclined towards realism about objects than you are. I find this curious because you and I both appreciate Gibson’s work on affordances, and I am a realist about affordances — affordances are real features of the organism-environment relationship But perhaps you are more inclined to think of affordances as “projections” from the organism onto the environment?
I’m not sure I am understanding the point there. Affordances are not objects. Moreover, what counts as an affordance will depend on the knowledge and interests of the person (the perceiver).
“Objects” are that which pushes back against us, thwarts us, offers resistance to our actions.
I’m not sure that’s completely satisfactory. But, ignoring that for the moment, even that conception of objects makes them pragmatic things rather than logical things. And it makes what counts as an object depend on our interactions with the world.
If Nature had no joints at which to be carved, then any criteria for successful action would be equally arbitrary as all other criteria.
I’m not sure that makes sense. The world is not homogeneous, so even arbitrary choices are not equally arbitrary.
I live in Illinois. The border of Illinois and Iowa is the Mississippi river. That could perhaps be considered a seam. But if I look at the border of Illinois and Wisconsin, part of the border is a river, and part of it isn’t (the river was not followed as far as it could be). The part that is not along a river doesn’t seem to fit the idea of seam. Even if there are seams, we are not bound to follow them and often don’t. And where there are no seams, we still carve up the world. So whether or not there are seams isn’t all that important (in my opinion).
If we were trying to define the boundary of Illinois and Iowa today, we would probably do that in terms of GPS coordinates, rather than using the river. The way that we divide the world depends on our abilities.
And here’s where I largely disagree:
So while we should always be on guard against the assumption that any theory correctly describes the structure of reality, we really cannot do away with the assumption that reality does indeed have an intelligible structure, and indeed one that is knowable by us because our cognitive capacities are a part of that structure and informed by its history.
If I climb a rock, I look for footholds. I don’t doubt that the footholds are real enough to support my weight. But the footholds are not part of the structure of the rock, they are just accidental inhomogeneities. That they are footholds derives from my pragmatic choices, from my temporizing. If the rock were completely smooth, I would not be able to climb it (or maybe I could find some glue pads, and then climb with those).
I see nature as like that, in that it is not homogeneous. So we can find something like footholds that we can use to anchor our descriptions. But that does not make those anchor points part of the structure of reality.
Here’s where I disagree with Gibson. According to Gibson, we pickup information from the immediate environment. I used to think that way, but it doesn’t work. The better view is that the environment is devoid of information. We create information. Part of what the brain does is creating information. We (or our perceptual systems) use whatever “footholds” or “anchor points” that we can find to anchor some sort of practical coordinate system to reality. And then the mathematics that we see in science comes from the very mathematical way that we have used those coordinate systems to systematically divide up the world.
When I look at scientific laws, a large part of these laws have to do with anchoring our coordinate systems and using those anchored systems to allow us to make measurements (making a measurement creates Shannon information).
When I look at Maxwell’s equations, they allow one to derive the wave equation. From that, we can conclude that our measurements of electromagnetic phenomena satisfy the wave equation. I don’t think it means anything to say that light itself satisfies the wave equation.
So now think of water waves. If I’m right about information, then perception works by virtue of our brains doing something similar to measurement. So when we see water waves, we are really seeing waves in our measurements. And, indeed, it is known that the water molecules just go up and down, or in small approximately circular motions. The molecules don’t advance in the wave (except when the wave is breaking near the shore). So perception is really looking at something like measurements that our brain is making.
So here’s my objection to phenomenology. As I see it, the phenomena are created by the brain (I think some folk would agree with that), and so it isn’t that we are seeing phenomena. Rather, it is that we (or our brains) are creating phenomena as part of what seeing is. The basic underlying idea of measurement (which is really a kind of categorization) is something that can be studied when it is done publicly by scientists. So, at least in principle, we can understand how phenomena arise, and need not just take phenomena as a starting point.
When I go through all of that, I cannot find a standard whereby we can judge the correctness (or truth) of a scientific theory. We can judge how well it works, but pragmatism is a guide, not a standard. That’s why I disagree with talk about whether theories are correct. We should limit ourselves to talk about how well they work. We can then talk of the truth of the data, if it conforms to the standards set by the theory.
They certainly are!
I completely agree. This is not about certainty it’s about knowledge.
Alan Fox seems to think that knowledge (from revelation) is indistinguishable from delusion. I just can’t see how that view can be supported
If truth does not exist then you can’t know you exist. If fact you can’t know anything at all.
The acknowledgement that you in fact know you exist is a indirect concession that truth (ie the logos) exists.
peace
Rational nonrealism, also called instrumentalism, is a family of views that agree on two main points: (1) Science is an objectively rational enterprise, and conceptual relativism is false as a way of understanding the rationality of science. (2) Scientific theories do not give us true or approximately true pictures of the unobservable, theoretical entities and processes that are causally responsible for what we empirically observe, nor do the theoretical terms in those theories refer to actual entities in the theory-independent world (and even if, by accident, they do refer to existent entities, this would be utterly irrelevant for science). Rather, scientific theories are useful tools or instruments (thus the name instrumentalism) that help us accomplish certain things. But what, then, are the things that scientific theories help us to accomplish? Here, instrumentalists differ, and at least four answers have been given to this question.
– Philosophical Foundations for a Christian Worldview
5. The knowledge the realist is talking about is the lived and experienced unity of an intellect with an apprehended reality. This is why a realist philosophy has to do with the thing itself that is apprehended, and without which there would be no knowledge. Idealist philosophers, on the other hand, since they start from thought, quickly reach the point of choosing science or philosophy as their object. When an idealist genuinely thinks as an idealist, he perfectly embodies the essence of a “professor of philosophy,” whereas the realist, when he genuinely thinks as a realist, conforms himself to the authentic essence of a philosopher; for a philosopher talks about things, while a professor of philosophy talks about philosophy.
– Etienne Gilson
A Handbook for Beginning Realists
Some readers may decide to become realists.
Others may decide that they are not realists after all.
From what I’ve been reading that is a vain hope indeed. I’d sure like to know how you could be persuaded.
ETA: “Constructive empiricism opposes scientific realism…” – Wikipedia
It is inductively unattainable. Cromwell’s rule and all that.
Cromwell’s rule was short lived.
“I beseech you, in the bowels of Christ, think it possible that you may be mistaken.”
Read for yourself:
https://en.wikipedia.org/wiki/Oliver_Cromwell
Err, if you are saying that it is vain for me to hope for realism about unobservables AND constructive empiricism, well, I knew that.
My point in the post you link was that I see the arguments currently as a toss-up so I take realism based on hope, but if the reason could be used to persuade me that realism about unobserbables was the wrong approach, then I would change my mind.
Three sentences replying to a post where I asked for more than one or two. How droll.
If what you take out of your posting of a sequence of long, unadorned quotes is that some people will believe in realism and some won’t, then I can accept that conclusion.
In fact, I would generalize it to say some people will believe in one thing, and some will believe in another.
I would also say one sees a lot of that here at TSZ.
FWIW, I understand from Kaufman that Davidson argues that if there is no translatability, we are not justified in calling something a conceptual scheme.
I think I can agree with that and would extend the pragmatic justification to science.
But I understand pragmatic realism as saying the success of our actions in the world based on a conceptual scheme warrants us claiming that scheme tells us something true about reality. “True” in Peirce’s sense.
What I am struggling with is understanding the realism and truth of pragmatic realism compared to garden-variety metaphysical realism and correspondence truth.
It’s interesting in the comments that you at first think you might be an anti-realist, according to (Kaufman’s version) of the philosophical definition. But then you seem to say a better understanding of your view is that reality is there but either (1) it is ineffable or (2) requires a conceptual scheme to say meaningful things about.
I take the ineffable position to be one Kaufman claims is incoherent based on Davidson.
I take the scheme-required position to be the internal realism, which from later comments is the form of anti-realism Kaufman himself embraces. From the NDPR book review he links to (worth reading in its entirety):
For the internal realist, on the other hand, though the world is causally independent of the human mind, the structure of the world – the individuals, kinds, and categories of the world—is a function of the human mind.
I’ll give my take on correspondence.
In order to have a correspondence theory of truth, there needs to first be some correspondence between reality and our linguistic expressions.
I take science to be engaged in constructing such a correspondence. In some sense, a scientific theory establishes a correspondence. We can only have correspondence-truth after such a correspondence has been established. So we can only judge a scientific theory pragmatically, in terms of the utility of the correspondence that it establishes.
However, I think the correspondence theory of truth is flawed for other reasons. If we really pay attention to what people say, you will see that there are many correspondences, and often mutually contradictory ones. So that should leave truth as relative to the particular correspondence that is used.
We don’t actually see a lot of this relativism. That’s because people are very adept at recognizing which is the appropriate correspondence, based on the context of the discussion.
As an example, consider the colorization of old movies (originally filmed in black and white). Based on one view of correspondence, the colorized films should correspond better with reality, so should be more true. But the critics of colorization objected, and thought that colorizing made them false.
My understanding of Churchland’s view of neural maps and scientific progress is based on the following quote from one of the reviews I read
Second-Level Learning [is] described in terms of high-dimensional neuronal kinematics for cognitive activity in general and vector-processing dynamics for perceptual activity in particular, [and] yields, in a non-linguaformal and naturalistic format of overlapping conceptual maps, an epistemological framework for cognitive activity “that sustains correspondence theory of our actual representational success” and “a coherence theory of how our representations are evaluated, by us*, for their representational success or failure” (202). Because this kinematical and dynamical framework reveals “the brain’s capacity to get an increasingly broad and penetrating grip on the structure of the larger reality of which it is a part” (247-8). Churchland argues that it provides sufficient epistemological basis for an optimistic meta-induction capable of vindicating his long held but somewhat eccentric (given his recognition that even perception is theory-laden!) scientific realism.
I read this as saying the success of our neural maps provides a justification for scientific realism.
My concern is that if science can only happen by using cognitive technologies to complement neural pattern matching, then any claims about scientific realism and progress must include an analysis of those technologies. Which is what I take a lot of “useful” philosophy of science as doing already.
So I was objecting to any claim that the analysis of scientific realism could be completely replaced by neuroscientific analysis of how successful neural maps are produced.
———————-
* that “by us”, which I take as a quote from Churchland, shows that even he can slip into what looks like homunculism or perhaps the double subject fallacy (pdf)
Thanks for our thoughts. My thinking on some of your points:
1. I also understand correspondence as trying to find some type of mapping from reality to some representation humans use. But, as I said in previous posts, I don’t think propositional logical or predicate logic or ordinary language are the right ways to think about scientific representation.
It does seem that understanding “reference” does require a way of naturalizing representation which, by the way, is one thing I understanding Churchland to be trying to do.
2. With respect to truth, reference, and context: my understanding of philosophy of language is that it divides into pragmatics, semantics, syntax. The analysis of pragmatics includes the role of context in determining reference. So I don’t think that the fact that reference varies by context is a problem for reference in philosophy of language or that it leads to unaccounted-for relativity in the theory of reference.
3. On colorization of old movies: Wasn’t the controversy about colorization of old movies more an an aesthetic criticism or perhaps a concern with failing to preserve the directors vision? If so, it is more about being “true to the artist’s vision” which I would take as a different sense for “true” than true in terms of the movie corresponding to the reality it photographed.
Neither of those is quite right.
According to the dictionary, “ineffable” means cannot be expessed in words. But it is more a matter of “cannot be expressed with our current vocabulary, so let’s expand that vocabulary.”
For the conceptual scheme thingy, I’d like to give an example from mathematics. And the example is linear algebra.
One way of teaching linear algebra is as a vector space, with vectors as triples (x, y, z) coordinates (or similar for other dimensional spaces). Another way is to describe it more abstractly. We can still say a lot about linear algebra without a coordinate system. We can even develop a theory of coordinate systems and of inter-translatability to them. But linear algebra, taught that way, probably looks more abstract to the average student, though a pure mathematician prefers the version without coordinate systems.
I see a conceptual scheme as much like a coordinate system. You need it to get to specifics, such as individual facts. But there are things that can be said without a coordinate system or without a conceptual scheme. So a conceptual scheme is a system of pragmatic choices that allows one to have specifics such as facts.
My view of reality, is that we face it not just without a conceptual scheme, but with nothing to which a conceptual scheme could be applied. We have to first acquire abilities to interact with reality before we can begin to think in terms of conceptual schemes.
That might seem unworkable. But that’s really an artifact of the intellectualism of philosophy. If I want to look at it more pragmatically, then we are animals who learn to interact with our world. We use a stimulus-response method. We stimulate reality and see how reality responds. We push and pull, and see what happens. It’s trial and error learning.
Language arises later, when we want to coordinate our pushing and pulling with that of others. And the excessive intellectualism of philosophy comes from trying to give a theoretical account of what language does.
In trying to understand human cognition, I looked at it from the perspective of what is required for a solitary animal to have a world. So I didn’t concern myself with language until I first had some sort of grasp of the requirements for an individual.
For example, I see ontology as pointless. To the philosopher, it seems to be a starting point — you have to know what you are talking about before you can begin to talk. But, for me, language is about cooperation, and ontology is an add-on needed to give an intellectualist account of language.
I gave the example of linear algebra above. The vector approach to linear algebra is what the engineer wants to solve his problems. And it can work without the abstract theory. The abstract theory is the pure mathematicians intellectualizing about the use of vectors.
So it similar with language. The ordinary use of language is the day-to-day use for cooperation, comparable the vector account of linear algebra. The philosopher’s intellectualist account is comparable to the pure mathematician’s abstract approach to linear algebra.
The one difference is that mathematics is always strictly rule-based. So the vector approach and the abstract approach both deal with the same linear algebra. In real life, the everyday use of language for cooperation doesn’t fit the more abstract intellectualist approach. So philosophers do a lot of idealization. And that idealization is where they can easily lose contact with how it all works. Language ain’t ideal, and an ideal language is probably unlearnable.
The reason that everyday language use doesn’t idealize, is because it is not rule-based. Wittgenstein’s argument on the impossibility of learning a rule seems appropriate here.
Yes, that’s also a good description of my view. But I want to start earlier than Kaufman.
In the interview, the telling difference was where Dan Tippens brought up the idea of carving up the world. And Kaufman objected. As best I can tell, he thinks you already need a structured world before you can begin carving. By contrast, I think you begin carving before there is any structure, and how you carve is part of how you create structure.
Yes, that seems right. But I don’t think that changes it relevance to what we mean by “correspondence”.
I understood Kaufman as saying the phrase “carving up the world” was already a concession to realism and he rejected that concession because it used “world” without modification by a conceptual scheme.
I take it as similar to him saying realists cannot even talk about the world without using a conceptual scheme, hence realism about the world independent of human schemes must be ill-founded (or incoherent or wrong or some negative adjective — I am not sure which).
I am not saying I agree with or even fully understanding the subtleties of these points.
That’s what you would get if you insist that language is primary.
If method is primary, then you start by engaging the world. You make up language later (which includes making up a word for the world).
I’m trying to catch up with this very interesting discussion, and I’m deeply grateful to both Neil and BruceS for having helped make TSZ a space in which in real philosophizing is possible. That is a rare thing indeed! Thank you both!
Now, to work.
Among the various issues that have caught my attention here, the one I find most intriguing is the question as to whether our use of pragmatic criteria for preferring one model over another — criteria such as explanatory unity, fruitfulness, promoting consilience, providing an alternative vocabulary in which anomalies are removed, and so on — in any depends on metaphysical realism.
Neil has given us a powerful set of arguments against realism. I think his position can be reconstructed as follows:
(1) any model of the world can only be tested by deciding which regularities to measure and which regularities to ignore;
(2) so, the tests or measurements are chosen based on the model we’re interested in testing;
(3) so, there are no model-independent measurements;
(4) but, the only intelligible sense to “how the world really is” requires that we can talk about some underlying real structure that the world has;
(5) so, talking about “how the world really is” requires that there is some underlying real structure independent of all model-dependent measurements;
(6) but, we have no mode of epistemic access to any such real structure, since all knowledge of reality is by means of model-dependent measurements;
(7) therefore, all talk of “how the world really is” is semantically empty and pragmatically useless — a stand-in or under-study for a God who has died.
I think that’s a really good argument — it makes really clear that the defender of metaphysical realism needs to show that we have some mode of epistemic access to the structure of reality independent of all model-dependent measurements.
I can think of pretty much just one way that the defender of metaphysical realism can respond to this challenge, and that is by transcendental argument.
The transcendental argument for metaphysical realism goes as follows: if it were not the case that the world had some structure independent of our model-dependent measurements, then we would have no non-arbitrary criteria for distinguishing between successful and failed test of models, because all tests and measurements would be equally mere ‘projections’ onto an undifferentiated and unstructured ‘reality’. But the history of knowledge, including scientific knowledge, gives us reasons for thinking that we do have at least some non-arbitrary criteria for distinguishing between successful and failed tests. So we have very good reasons for thinking that reality has some structure independent of our measurements.
However, it is one thing to establish that reality has some measurement-independent structure, and quite another to say anything about what that structure is.
The distinction I want to make here is between
weak metaphysical realism: we can know that the world has measurement-independent structure, but we cannot know what that structure is (because any description of it would be measurement-dependent);
strong metaphysical realism: we can know that the world has measurement-independent structure, and we know at least something about what that structure is (because we have some mode of epistemic access that doesn’t involve taking any measurements).
Putnam’s arguments for internal realism are arguments against what he calls “metaphysical realism,” but which is what I call “strong metaphysical realism”. Weak metaphysical realism is still an option.
In fact, I would even strengthen weak metaphysical realism slightly here — we are entitled to posit transcendentally just the minimal structure that reality must have in order for us to have testable models at all.
The real problem is, how does one distinguish genuine revelation from delusion? If I hear God’s voice, how do I know whether He is speaking to me or I’m having a schizophrenic episode? What are the criteria, and what makes those criteria reliable? Or is there a ‘leap of faith’ involved in taking what one experience as revelation to actually be revelation?
Firstly, I’d wouldn’t say that I know that I exist; I would say that I am certain that I exist, because I don’t understand what it would mean for me to deny my own existence. Descartes was right about that much, though mistaken to think that this certainty was a foundation for knowledge.
Secondly, just because I am always certain that am experiencing, and usually (but not always) certain of what I am experiencing, it doesn’t follow that I know that reality is precisely as I experience it as being. There is no direct route from phenomenology to metaphysics.
This seems confused to me in two ways.
Firstly, I don’t know what “truth exists” means. Truth isn’t an odd sort of entity, like a rock or a number, so I don’t know what it means to say that “truth exists”. Truth is, according to a correspondence theory of truth, a relation between a proposition and the aspect of reality that the proposition refers to.
Secondly, as I understand it, the Logos is the intelligible structure of reality. So in affirming the Logos, one affirms that reality is knowable. I don’t see any reason to believe that the intelligibility of reality only makes sense within a theistic framework. For one might think that it is cognition is a biological function, sculpted over millions of years of natural selection, and that the “fit” between cognition and environment is no more mysterious than the “fit” between a plant’s roots and the soil in which it is rooted.
A naturalistic theory of cognition is no refutation of theism, of course — but it is to say that we do not need to presuppose theism in order to do epistemology at all.
I think that’s too strong a claim. And there’s the additional difficulty that it is far from clear what it means to say that the world has structure (as distinct from saying that we experience some structure).
We can, of course, say that the world is structurable. But I think that’s too weak, since modeling clay is structurable.
We have plenty of evidence that we are able to discriminate between different parts of the world. But that isn’t really a property of the world, because it takes us or some other agent or organism to do the discriminating. And it isn’t really a property of us, either. I’d call it a relational property on how we relate to the world. But it differs between people. A color blind person discriminates differently from a person with normal vision.
I’ll add that the question of what we can say about the world is not one that I see as important. For me, the bigger question is what can we say about humans, that they are able to cognize the world.
Do you think it is impossible for God to reveal something to me in a way that I can be sure it is not a delusion?
You do know that there have been several criticisms offered to Cogito ergo sum
Are you sure that you have adequately defeated every possible criticism to date and can defeat every one that could ever be offered?
I’m not at this point making any theistic claim. Although I might later if I have time. What is important right now it that you accept the necessity of the Logos for knowledge
peace
Are you saying that agents are properly not part of the world? What makes agents and their properties, such as ability to discriminate, stand outside?
I’ve tried before on this, but since KN is now using the term, I’d like to try again to understand your use of the term “measure”.
In the OP, you say “perception works by virtue of our brains doing something similar to measurement”. Above, you talk about discrimination. How do the two relate? What does “measure” mean to you?
A dictionary definition of “measure” would be roughly assigning a number in some standardized way to an entity. But I assume your use is inspired by measure theory in math, which I understand as (roughly) a formalized way to assign a real to subsets of some specified set. The subsets are taken as a given, I believe.
I would say that the forming of the subsets is discrimination and it must precede (both temporally and explanatorily) the forming of the measure. But by having measure part of perception, you seem to combine them.
No, I am not saying that. You are reading too much into the particular wording.
I’m curious as to what you mean by “measurement”. To me, it has to involve assigning a number; further, the entity to which the number is to be assigned is taken as a given.
In your paper, you talk about a “measurement” stance with respect to science. I can see how it makes sense to talk about assigning numbers for the context of scientific models and theories.
But here you seem to be using measurement/assigning numbers as part of what any creature does in interacting with the world. But I would have thought perception would precede that. Further, if we assume perception is direct, as I believe most pragmatists do (including Putnam these days), then I would understand that as already providing a structure for the world (based on the creature’s evolutionary history and possibly the application of learning mechanisms so created).
I’d also wonder how this whole issue relates to the lower level branch of your bifurcated intentionality, ie the sensorimotor ways of interacting with the world which involve norms for success. Again, this seems to presuppose the structuring into objects exists (at a personal/whole creature level, of course).
(In response to your thank you, let me repeat how valuable it is to me to be able to interact with the trained philosophers on TSZ, and in particular you and Walto).
Let’s try again.
If not property of the world, then property of what is the ability to discriminate? Yes, I know what immediately follows – it’s a property of agents. But agents are part of the world, aren’t they? Yet you are saying that ability to discriminate is not really a property of the world, even though agents are part of the world. How does this add up?
Okay.
I’ve looked for philosophical treatments of measurement, but I haven’t found anything useful.
So here’s my analysis.
We carve the world up. Philosophers generally agree about this. I’m using “categorization” to refer to that.
Measurement amounts to:
1: Carving the world up into very thin slices, in a highly systematic manner (such as comes from geometry).
2: Giving those slices numeric names, again in a highly systematic manner, so that nearby slices have numeric numbers with small differences. That way, the mathematical relations of the numbers reflects the adjacency relations of the slices.
This is how we impose mathematical structure on reality. And this is why nominalism (numbers are just names) is entirely adequate to account for the usefulness of mathematics.
In practice, we find anchor points, and interpolate our divisions (or carvings) between those. For example we use the freezing point of water and the boiling point of water as two anchor points for temperature measurement. The other divisions are interpolated between those.
It is known that the brain contains arrays of very similar neural “devices”. I assume that these are for doing something similar. It is easiest to detect edges (say the edges of an object). Presumably the brain is interpolating division points between those easy edges. That’s what allows us to scan across an apparently uniform surface.
Note that I am not an experimentalist. So the above paragraph is mostly guessing.
While the brain is likely doing the same sort of interpolated carving up, I doubt that it is naming the slices with numeric labels. I doubt that it is naming them at all. I expect that the activated neurons are being used in place of a name.
As for perceptual discrimination, that would leave the primary discrimination at the anchor points (the edges of an object), with a secondary discrimination due to the interpolation of additional ways of carving up. But it all depends on the anchor points.
As an aside, I’m hesitant to fully embrace big bang cosmology, because it lacks the anchor points. It is mostly extrapolation, based on what we know about the local region. And extrapolation is far more risky than interpolation.
Actually, no. It’s inspired by how I see us doing ordinary measurement.
I’m treating science as perception written large. Science uses the best means it can to get data, and perception faces similar problems so is likely to be using similar methods.
I can put measure theory into perspective. It’s an idealization of how we measure area (or volume). Imagine a farmer planting his crop. But his field has clumps of trees, some creeks, and a few bogs or marshes. So where he plants is scattered around. We would like to be able to talk of the area used for crops. So that’s the area of an irregularly shaped subset of the field between the obstacles. Measure theory idealizes that idea of determining area (or measure of subsets).
I hope I have clarified how I am using “measure.”
But that’s unhelpful.
If you want to say that we are part of the world, then fine. We should stop doing philosophy or science, or anything much. We are just part of the world.
But if you want to think about the world, then you have to adopt a stance whereby you hold yourself separate from the world. Philosophy and science are all done from such a stance. Yes, we should recognize that it is only a stance. But it is a stance that informs our discussions of these topics.
Not without violating your free will, I think.
fG
BruceS,
I think I probably did equivocate between measurement and perception, so thank you for bringing that to my attention.
I do think of perception as more “basic” than measurement. I am a direct realist about perception, though more in line with J. J. Gibson (and before him, Dewey) than with Putnam. Putnam’s direct realism is taken straight from McDowell, and I think of that as a step in the right direction but ultimately too thin and too formal to be correct.
What you don’t find in McDowell, or in Putnam, is any account of the historical, cultural, political, and economic forces that sculpt the formation of our perceptual and conceptual capacities. In veering away from the excessive constructivism of Rorty, they have forgotten the whole line of thought inaugurated by Hegel, Dewey, and the Frankfurt School to show that constructivism and realism are compatible.
A further distinction would need to be drawn here between biological constructivism — that what an animal perceives of its environment is partly specified by its perceptual and conceptual abilities — and socio-cultural constructivism — that what an enculturated animal perceives of its environment is partly specified by how its cultural inheritance modifies its perceptual and conceptual abilities.
But I think that both biological constructivism and socio-cultural constructivism are compatible with realism, as Dewey tirelessly (and tiresomely) showed in Experience and Nature and Logic: The Theory of Inquiry, just because experience is both in nature and of nature. For experiencing is what animals do, and animals (including rational animals) are a part of nature and its history.
I think that is a bit too strong, at any rate for my tastes.
The stance we adopt in doing philosophy and science is that of trying to solve a problem, and the recognition of a problem as a problem is an interruption in our absorbed coping with the causal powers that populate our environments. The problem-solving stance does involve constructing a model of the situation in which one finds oneself, and the manipulation of the model is how we discover the solution (or a family of solutions).
Put that way, any problem-solving animal engages in that kind of cognitive activity. My cat did that all the time when I first got him, and he had to determine if he could make the jump from one bookcase to the next. By now the layout of the furniture is routine to him — the living room no longer confronts him as consisting of problematic situations.
What makes philosophy and science distinct is that we can collaborate in constructing models, offer corrections to each other as to the adequacy of our models, and offer corrections to each other as to whether our solutions are optimal or sub-optimal. Doing all that requires culture, included a shared natural language. I think that writing plays an important role here as well; I don’t think it’s a coincidence that philosophy emerges in every human culture shortly after the invention of writing.
In any event, we can make sense of our ability to do philosophy and science in these terms without invoking the rhetoric of “holding oneself separate from the world.” I worry that such rhetoric invites all sorts of Cartesian anxieties that are better exorcised than answered.
Consider it this way:
(1) I experience a divine presence because God is revealing Himself to me;
(2) I experience a divine presence because I am having a schizophrenic episode.
If the two experiences have identical phenomenology, then no appeal to the phenomenology can tell me whether (1) or (2) is the case.
The only major criticism of the cogito ergo sum I know of is that Descartes is not entitled to infer, from the fact of thinking, to the postulation of a being that does the thinking. Put otherwise, the event (or process) of thinking is too “thin” to ground the thing-property distinction that he goes on to use.
There are other criticisms of the various Cartesian circles that Descartes gets himself trapped within, when he tries to use the cogito to ground all other knowledge-claims, but I’m not interested in defending any of that. (I consider Descartes to be a massive blunder for Western philosophy.)
The only point I was making there is that the fact of one’s own experience is unquestionable to the person whose experience it is.
I don’t recognize my own views under that description, but perhaps its not crucial that I do.
I do think that the possibility of knowledge requires recognizing that the world has a structure unto itself with regard to its properties and relations that are the causal ground of the regularities and irregularities in our sensations. If you want to call that “the Logos,” that’s fine by me — but that’s not a term I would use.
Thanks for the detailed response, Neil.
Of course, there is no need to actually do experiments in order to engage with the experiments of neuroscience (a fact that many philosophical careers attest to!).
Neuroscience claims many conclusions about how neural processing works to guide and learn from bodily interactions with the world. So it would seem it is possible to see how your theory fits current neuroscience without doing experiments yourself, although of course that would be substantial effort which may not be of interest to you.
I read Rosenburg’s 1988 paper on convergent realism. I agree that his goal is to show that theories converge and, importantly, do so in a way that avoids the issue of incommensurability between the terms used in a successor and predecessor theory.
He claims early in the paper that he will do this using an analog of mathematical Cauchy sequences; to show convergence under this analog would require that the “distance between theories” can be made as small as asked by going far enough in the sequence of theories. As far as I can see, he never really addresses the distances between theories getting smaller. Instead, the paper is about how how to show a successor theory makes progress in a way that avoids the issue of Kuhnian incommensurability, as you say.
But that’s a side point; the main issue here is realism. He says that we do science for the purpose of telling “a story of how things really are” (paraphrased lightly). For me, the key paragraph on this occurs new the end of the paper:
“What makes the central outstanding question for our inquiry is the connection between the “convergence” picture of scientific progress and realism. For according to the pragmatist ontology which informs this normative epistemology for physical science, “the world” just is whatever would be represented as existing at the limit of such a convergent sequence of theories. More precisely, talk about “the world” simply amounts to talk about this temporal sequence of world-stories, and, in particular, to say that there is or exists only one world just is to say that the diachronic sequence of pairs of theories related as predecessor and epistemically-qualified successor does “converge to a limit”.
That seems to be aligned with what you say.
But for me, it leaves this question open: is the realism he promotes defining reality as whatever the world stories converge to? I would take that as a form of constructivism.
Or is he saying that the sequence of theories converging means there is an metaphysically objective reality and the converging sequence of world theories tells us more and more about its nature.
You say in another post that it can be both at the same time; I’m not clear on how that could be.
BruceS,
I think it would require explicitly testing the hypothesis that something corresponding to measuring is going on. If the researchers are looking only for computation, they probably won’t find measurement.
I need to try some Dewey in the original at some point; for now, I’m sticking with secondary sources.
For his views on realism, I found a paper (pdf) by Peter Godfrey-Smith which I read as interpreting Dewey as a ordinary-object and scientific realist, with the qualification that Dewey did not want to privilege intrinsic properties over relations. PGS also says Dewey used the words “mind-independent” and “instrumentalism” differently than we do now, so one has to be careful in reading him.
I have had some extensive conversations with people having delusions — people undergoing court ordered medical intervention. I do not think it is possible to be certain that we are not delusional. There is no separation between what we judge with and what is broken.
Of course I could be deluded about this.
“Computation” is a word that is used in many different senses by experimenters and theorists, let alone philosophers: GOFAI/LOT, connectionist, probabilistic, control theory, and many more, I suspect.
So I think it would be hard to generalize without looking. Lots of work to do so, of course.
ETA: Plus, computational modelling is used to provide cogent explanations. What neural networks are “really” doing is biochemical. So it is a matter of mapping a computational model to a biological process and structure in a way that explains and predicts best.
There’s a big problem with this way of thinking.
The problem is that, by itself, a “world story” is merely ink marks on paper. For it to be more than that requires that meanings be independent of the world story that they are presenting. And I doubt that is possible.
Take, for example, the transition from Newtonian mechanics to Relativistic mechanics. People who are not well acquainted with the physics see this as a huge change, with one contradicting the other. Physicists, on the other hand, are more likely to factor in the change of meanings between those appropriate for Newtonian science and those appropriate for relativity. And when one factors in those changes of meaning, the changes between Newtonian science and Relativity seem far smaller.
Convergent realism would need to be investigated and discussed in a completely neutral objective language. And we do not have an candidate for such a language.
This might be why I tend to see myself as a realist, while KN sees me as anti-realist. One of my big concerns is the lack of a neutral language. And part of what I see science doing, is building an adequate language for the science of the day. When I deny that the world has structure, it is mainly because I do not know of a neutral way of defining structure. We certainly see structure according to our current meanings, but those meanings are not neutral.
In an earlier reply, I tried to describe in as close to a neutral language as possible. That’s where I described science as carving up the world, and I suggested that scientific progress can be seen in how finely we carve up the world.
What Kuhn would have described as normal science is something like continual improvement in our measuring techniques, both to increase the precision of our measurements and to extend them to the more distant reaches of the world. Revolutionary science is where we reach a point that we realize that there are limitations in how far we can extend our measuring system without a major realignment.
As a simple example, people assuming a flat earth could measure distances with a north/south/east/west grid. But once you realize that the world is closer to being a sphere, you see that such a grid cannot extend to all of the earth. So you need a major realignment in your measuring system (and in your system of meanings) to accommodate that sphere-like earth.
I’m curious what computational model best aligns with neuronal behavior and brains.
Rosenberg’s paper is about the problem you describe of relating the meanings of incommensurable theories.
But he does not solve that problem by proposing a neutral language. Instead, he tries to show how a successor theory can explain the successes and limitations of the predecessor, and so be an improvement, while still only using the language of the successor theory.
He claims this can be done by creating a counter part theory of the predecessor theory in the successor by embedding that older theory in the new one as a counterpart which is a “nomological isomorphism” of the predecessor.
He contrasts mapping of the old theory to a counterpart in the new theory with logical positivist accounts of theory progression which depend on logical deduction of predecessor from successor and so require common terms or a neutral language in which both can be phrased.
However, he only gives one toy example. Also, according to cites on philpapers, no on else has built on his work.
I think there’s a problem in that successor theories can be unnecessarily cumbersome. No one needs general relativity in everyday life, even if you are calculating trajectories or orbits.
A problem is that no one seems to have taken on Rosenberg’s 1988 ideas to help detail them and check them against many examples of theory change. In philpapers I see only one cite to this paper, and that is by Rosenberg himself (however, I don’t know if that philpapers mechanism is reliable — it is in Beta).
However, Rosenberg says only theories that can produce quantitative observations can be checked because doing so requires looking either at the equations themselves or at the fit of equations to data. So it might be that using his technique of “nomological isomorphism” would involve mapping the mathematical structure of the old theory to a counterpart in the new one. In the paper’s toy example of the theory of gases versus molecular kinetics, he does just that.
So perhaps a detailed version of Rosenberg’s work would turn out to be a form of scientific realism via structure.
Come to think of it, Rosenberg’s 1988 paper does pre-date Worrall’s 1989 revival of scientific realism via structuralism. So maybe that is why he did not call it structural realism!
One way to think about metaphysical realism and conceptual schemes is to ask whether content can be separated from scheme. In other words, is there a content of the world that schemes can track, or is any discussion of content inseparable from scheme.
The SEP article on challenges to metaphysical realism states
On all fronts, debate between realists and their anti-realist opponents is still very much open (Khlentzos 2004). If realists could provide a plausible theory about how correspondences between mental symbols and the items in the world to which they refer might be set up, many of these challenges could be met. Alternatively, if they could explain how, consistently with our knowledge of a mind-independent world, no such correspondences are required to begin with, many of the anti-realist objections would fall away as irrelevant.
There is talk of pragmatic realism in this thread, but here is an article on pragmatic theories of content, which has a new idea to me.
Manjaro.