I hope I will be forgiven for abusing the term “skepticism” here — for what I have in mind is not a perfectly innocuous “claims require evidence” epistemic prudence, but rather Cartesian skepticism.
According to the Cartesian skeptic, one can be perfectly certain about one’s own mental contents and yet also be in total doubt about what really corresponds to those mental contents. Hence she needs an argument that will justify her belief that there is any external reality at all, and that at least some of her mental contents can correspond to it.
There are many responses to Cartesian skepticism, and here I want to pick up on one strand in the pragmatist tradition that, on my view, cuts deepest into what is wrong with Cartesian skepticism.
I think that one cannot talk, in any intelligible sense, about justification in the first place without also committing oneself to a belief in other minds with whom one shares a world. (Not that I like that way of putting it — “a belief in other minds” is a much too intellectualistic interpretation of the myriad ways in which we experience the sentience of nonhuman animals and the sentience-and-sapience of other human animals.)
I say this because justification is itself a social practice — and one that we ourselves are taught how to participate in. (In the contemporary jargon, I’m a social externalist about justification.) For what is justification? It is a normative assessment of the evidence and reasons for one’s claims. But that normative assessment necessarily involves other rational beings like ourselves.
Think of it this way (taking an example from Wittgenstein): suppose I’m waiting for a train, and I want to know if it will be on time. I could look up the schedule. But suppose further that instead of doing so, I imagine the schedule: I look up the time in my imagination. Why isn’t that the same thing as looking up the actual schedule?
The answer is that there’s no constraint on how I imagine the schedule. It could be whatever I want — or subconsciously desire — it to be. But without constraints, there are no norms or rules at all.
Justification is much the same: it is a normative assessment of evidence and reasoning according to rules or norms, and there are no private norms. (Though Wittgenstein doesn’t put it this way, he might say that the very idea of a “private norm” is a category mistake — a category mistake on which Cartesian skepticism and several hundred years of subsequent philosophy have depended.)
So whereas the Cartesian skeptic thinks that we need to justify our belief in the world and in other minds, I think that this makes no sense at all. We cannot justify our belief in other minds and in the world because there is no such thing as justification at all in the first place without also accepting (what is indeed a manifest reality to everyone who is not a schizophrenic or on a bad acid trip) that there are other sentient-and-sapient beings other than oneself with whom one shares a world.
.A further point to make (and the subject of my current article-in-progress) is that justification and truth require both sentience and sapience.
The clue I’m following is Davidson’s triangulation argument: suppose there are two creatures who are each responding sensorily to some object in a shared environment. How is an onlooker supposed to know which object they are both responding to? If both creatures can compare its own responses with the responses of the other creature, then each can determine whether or not they are cognizing the same object.
The point here is that two (or more) sentient creatures — intentional beings that can successfully navigate their environments — can each have a grasp of objectivity if and only if each creature can
(1) represent the similarities and differences between its own embodied perspective and an embodied perspective occupied by another creature and
(2) be motivated to minimize discrepancies and eliminate incompatibilities between its own action-guiding representations and its action-guiding representations of the other creature’s action-guiding representations, and in the process
(3) attain the metacognitive awareness whereby it can take its own embodied perspective as an embodied perspective, and thereby be aware that how it subjectively takes things to be is not (necessarily) how things really are.
This process is facilitated by a shared language that allows each creature to monitor how each is representing the other’s representations and revise its own representations when incompatibility between representations is discovered. The function of norms — of discourse and of conduct — is to motivate each creature to revise its representations when incompatibilities are discovered.
One important implication of this argument is that sentient creatures cannot distinguish between their own subjective orientation on things and how things really are. They lack an awareness of objectivity and an awareness of their own subjectivity. By contrast, sapient creatures are aware of both objectivity — how things really are, as distinct from how they are taken to be — and subjectivity — how things are taken to be, as distinct from how they really are.
This line of thought also explains why I have been adamant that objectivity does not require absoluteness: sapient creatures can be aware of the difference between how things are and how they are taken to be, and thus be aware that they might have false beliefs, even though no sapient creature can transcend the biological constraints of its form of sentience.
Seriously, I think petrushka has hit on the real issue.
Modernism thought it was making God unnecessary but what it was in fact doing was undermining the very foundation that made rational thought possible in the first place.
We still get along pretty well until we take stop to examine the edifice on which it is all built and find it to be nothing but sand.
Some folks like KN claim to be fine with that conclusion.
They even go so far as to claim that shifting sand is all there is while standing on nothing but shifting sand.
peace
You haven’t noticed that no one is impressed?
Fifth, it is the simplest thing in the world to build self-consistent word castles.
Doesn’t mean anything if it doesn’t connect to reality.
Oh I’ve noticed.
That response is easy offering a workable alternative is hard apparently.
peace
You say that but fail of offer a consistent set of axioms with which to build a worldview
again you fail to demonstrate how my worldview does not connect with reality.
You fail to even demonstrate how you could possibly demonstrate this given your starting presuppositions.
peace
The book he and De Caro published last month has three essays on perception plus one on Boyd (which I hope will speak to Boyd’s approach to (scientific) reference). My library has a copy of that book on order; I have not seen it yet..
If I’m allowed to conflate reference and intentionality, there was this from a post of yours from last November (it stuck in my mind because of its “I wonder what he meant by that” quality). Was there a specific Putnam publication you were thinking of?
Axiomatic logic is computer logic. It is what I was talking about when I used the phrase computer envy.
The world and its relationships are fuzzy. It has taken hundreds of years for science to tease out reliable relationships.
In case you have not been following Brain’s blog, this series of posts addresses some things that we’ve covered in this thread (epistemic access, embodiment, metaphysics) in combination with at least one we have not (limits “imposed” on intelligence by the theory of computation).
However, the author’s book looks to be beyond my pay grade in at least two ways.
Is this claim of yours fuzzy or concrete? Do you even bother to examine what you beleive and the implications before you post it as gospel?
This thread is still more evidence that you don’t even have a basis for concluding that the simple straightforward findings of science are valid.
How can you possibly “rely” on supposed findings that took hundreds of years to tease out.
peace
Have you ever been to a doctor or dentist? Do you own or us a computer?
You still don’t get. I can trust science because I can justify my knowledge given my worldview.
For you on the other hand these things must be just like magic.
peace
Whatever floats your boat.
How’s your programming project coming along?
Here’s the thing: there’s an established philosophical discourse that you seem not to be interested in using, and I am.
When you say “[w]hat makes a belief justified is whether or not it corresponds more or less closely to observed reality”, I want to say, “no — that’s what makes an empirical belief (and not all beliefs are empirical!) true.” Justification and truth are quite distinct concepts, and that’s not my arbitrary stipulation — that’s central to Western philosophical discourse since Plato (in the Meno, if you care to know).
Rather, in that tradition, justification is centrally tied to reasons. My belief that fracking would be bad for our community is justified if I can give the right sorts of reasons — some of which will be factual (e.g. probability of water contamination) and some of which will be moral (e.g. that we should do more to reduce fossil fuel use). My belief that fracking would be good for our community would be unjustified if I held stock in the fracking company — for even if fracking were good for my self-interest, that’s not the right reason for thinking that it is good for the community as a whole.
I tie the concept of justification to the concept of being justified in one’s beliefs and of being able to offer justifications for one’s beliefs. Those are deeply connected with reasons. Evidence is the right kind of reason for empirical beliefs — beliefs about the observable, actual world — but of course not all beliefs are like that. We have beliefs about a priori domains logic and mathematics, and many philosophers (though not myself) have thought that morality is a priori. In a priori domains we are not looking for evidence that makes one belief more likely than another, but for proofs.
My belief that fracking is harmless to human health can be perfectly justified if I’ve done everything I reasonably can to find evidence for and against that belief and the evidence available to me shows that it is probably harmless. But that doesn’t make the belief true. And I can have plenty of beliefs that are true even if I can’t adequately explain why I hold that belief.
Although I perfectly agree that nonhuman animals can often hone their mental models through trial-and-error, nevertheless
(1) nonhuman animals do not engage in a shared practice of give-and-take between their mental models in order to come to an agreement about what they ought to do collectively, and
(2) nonhuman animals do not deliberately test their models against the world in order to construct models of how the world’s hidden causal structures generate detectable affordances.
which is to say that nonhuman animals lack both justification (1) and truth (2), and so their reliable representations of their environments fail to count as justified true beliefs.
As I indicated, their knowledge is not justified because it is not connected to the social practice of giving and exchanging and evaluating reasons.
Here’s another way of putting our disagreement. You want to treat “justified” as equivalent to “reliable”. I dispute that, because I treat “justified” in terms of “reasonable,” which in turn means having reasons.
If we were to deny the distinction between (1) sentient knowledge: having highly reliable, affordance-detecting, action-guiding representations of detectable and often motivationally salient features of the local environment and (2) sapient knowledge: having sufficiently good-enough reasons for one’s true beliefs, then we would have to say all of the following
1. A spider has good reasons for believing that there is an insect in its web.
2. A tadpole has good reasons for believing that there is algae nearby.
3. A trout has good reasons for believing that there is a swarm of gnats above the surface.
4. A crow has good reasons for believing that if it were to make a tool of the correct shape, the food could be retrieved.
5. A chimpanzee has good reasons for believing that if it were to arrange the nut on a sufficiently flat stone and hit with enough force with a sufficiently sharp rock, the nut would crack open.
I am open to the possibility of accepting (4) and (5), though even here I’d need more argument than has been marshaled so far. (1)-(3) do not seem like viable options to me — they seem like attributing far more cognitive complexity than necessary to explain the observed behavior.
I have been following those posts. However, his mention of computational complexity suggests to me that he is on a wrong track.
I think you’re right. Part of what I’m trying to determine from this conversation with KN is what exactly he means by justification and how justified knowledge differs operationally from empirically tested knowledge. I suspect it may be down to mere definitions at the end.
I’d say that you are the one who doesn’t get it — particularly if you are depending on your worldview to justify your knowledge of science.
BruceS,
Yes, I’ve been following that — very exciting work!
He doesn’t need to provide an alternative in order to demonstrate that your claims are nonsensical.
Also not his job. You have the burden of showing that it does. Thus far you have not shouldered that burden.
Yes–Reason, Truth and History. Maybe his greatest work.
Empirical testing is justification, for the actual world insofar as we can causally interact with it with our senses or instruments.
Empirical testing would not justify an a priori belief. That’s why empirical testing is a species of justification, and not the whole of it. And what’s why we need a theory of justification that does not depend on the specifics about empirical testing per se. My account in terms of justification as a social practice of reason-giving does that, by showing that even what counts as “evidence” depends on the socially instituted criteria for evidence.
This makes justification a local matter. It depends on what counts as evidence in that time & place. It used to be thought that witches were often disguised as rabbits, and that if one had a piece of a witch on your person, then you would be protected from witches. That’s why it is thought that a rabbit’s foot was good luck. This belief was justified in light of beliefs about witches and the prevailing standard of evidence. It was quite false. But nevertheless justified.
This is not a disagreement about arbitrarily stipulated definitions. This is a disagreement about whether the Western philosophical tradition has any authority for us over how we use terms like “justification” and “truth”.
Is our use of terms like “justification” and “truth” going to be at all indebted to the works of Plato, Aristotle, Descartes, Hume, Kant, and Mill — or not?
‘
Well, no, it’s to say that they lack language, and so do not think in terms of justification or truth.
We get it, language is important.
Justification often must end up doing what? Showing something, or running through an experience. It comes down to sensory experience, which pretty much underlies everything we know about stuff outside of our heads (arguably, also what’s inside of our heads, but that needn’t detain us). Naturally, language makes this much richer cognitively, however knowledge about the world is based upon sensory experience, language or no language. “Justification” is a way of referring back to sensory data and of coming to conclusions from that.
That there is a difference between propositions and justification built upon empiricism on the one hand and animal knowledge on the other hand has never been in question. But that we know anything through the more basic processes of perception, of sensory experience, is also the case. Language allows us a far more elaborate understanding, but it still goes back to what we “see,” which is what “justification” is about.
In the end, it’s about what we can know based on what we see and can rationally glean from that. Justification using rationality is a wonderful tool for expanding knowledge based on that, the point is that justification is built on what we know via our senses and can figure out from that. Somehow it seems to come down yet again to the fact that we’re verbal and other animals are not, meaning that we can rationally process knowledge but must justify it in order to do so. Yet it always comes down to basing it on non-verbal sensory data (and agreed-upon rules for understanding those, of course), and justified knowledge is abstracted from that.
Language means that we can talk about knowledge, while justification is needed to make sure that we’re not just discussing fiction. It’s important, but ultimately it’s really about sensory data, which also provides knowledge to the non-verbal animal. If there’s more to it than that, it would needs some sort of basis like FMM rather unconvincingly claims to have.
Glen Davidson
My own opinion — not very popular here — is that pre-science philosophy and theology are much like alchemy and astrology. They represented the best thinking in their time, but are no longer producing new and useful ideas.
You mean he fails to make the same mistake?
Why should he create a fiction given ad hoc fictional powers that would address the problem–if those fictional powers were real?
That’s your mistake, and a fatal error in that we are provided no reason to think that it is not mere fiction.
Glen Davidson
Surely much of the logic remains sound today. But of course even that has been bettered more recently.
But sure, I don’t know why I should care about what Plato wrote, except for its historic influence. Indeed, his is a virtually religious view of language, logic, and thought, and I hardly feel beholden to what he wrote.
Dead hand of history, indeed.
Glen Davidson
I’m happy to use terms that we mutually agree on. I am not a professional philosopher, hence my questions about what you mean by “justified”, among other words.
That’s fine, but it does mean that you are defining “justification” to require language.
“It works” is a perfectly valid reason for holding a particular belief.
You seem to recognize here that being justified and being able to offer justifications are not the same. This addresses, I think, my confusion about your position. It seems to me that you are saying that unless one can use language to explicate one’s reasons for one’s beliefs, the word “justified” cannot be applied to those beliefs. Again, a matter of definition.
Does your definition of “justification” require this?
I’m not sure this is true. Some non-human animals do try multiple approaches to achieving their goals and appear to learn from those that work.
They may lack philosophical justification by the definition that you are using, but the fact that they learn behaviors that work in the real world suggests that they have arrived at a true (more or less) mental model.
In the absence of a social group, how would you operationally distinguish between justified and unjustified knowledge?
Again, “it works” is a great reason to hold a belief.
I think all of those situations represent good reason to hold a particular mental model, within the capabilities of the physical brain holding that model.
This I disagree with. What counts as evidence is not merely socially constructed, it must reflect feedback from reality.
I appreciate your pointers and respect your desire to discuss in that context. I’m happy to use those definitions in this thread. I disagree that the concept of empirical truth requires any context other than reality.
I assume sound logic remains sound logic.
What I am trying to say is that logic is largely irrelevant.
What produces the advances we see in scientific understanding is more like evolution. We try things and keep the things that work. We try to find reliable relationships, but this is more nearly rationalization than reasoning. More like curve fitting.
I would put it something like this: the criteria for what counts as evidence are grounded in social practices, but whether those criteria are actually satisfied depends on whether the world cooperates. I do think that the world gets a vote in what we say about it — I’m not Rorty!
I would say that empirical knowledge requires the cooperation of social norms insofar as they affect our conceptual frameworks and the world’s causal structures insofar as they affect our sensory receptors.
Let me take up my position here from a different angle.
Andy Clark (and other philosophers of cognitive science) argue that brains are constantly generating predictions. These predictions are affordance-detecting, action-guiding representations of some features of the environment that are detectable and often (not necessarily always) motivationally salient. The information being propagated from sensory receptors to the brain conveys the degree to which the predictions are wrong.
The basic cognitive architecture, from a neurocomputational perspective, is a “bidirectional multilevel hierarchy” (I would prefer “heterarchy” but that’s a slight quibble.) That is, information is being propagated across and within neuronal assemblies from the peripheral sensory receptors to the CNS (and back), as well as to motor control units. The top-down direction generates predictions and transmits them to the sensory receptors; the bottom-up direction generates prediction errors and transmit them to the CNS for the production of the next cycle of predictions.
On this picture of animal cognition — informed both pragmatist philosophy and contemporary neurcomputationalism — there’s no such thing as “sensory data” in the empiricist sense. (This goes back to my long-running argument with Erik.) Instead, animal brains are continually navigating the top-down and bottom-up representations in order to minimize predictive error while satisfying the animal’s needs and goals.
The difference that language makes is quite profound, because it allows from a whole new dimension of information transfer: top-to-top. Instead of training an animal what to do, continually reinforcing the behaviors you want and curbing the behaviors you don’t want, you can teach someone by telling them what to do (and what will happen if they don’t). This allows us to engage in linguistically mediated shared behaviors (“on the count of three, lift!”), because the top-to-top information then propagates downwards to the motor units of each individual person.
That’s why I think language makes a more significant difference to animal cognition & knowledge than others here seem to think.
Have you talked to a creationist recently?
What creationists count as evidence is socially constructed.
What Dubya counted as evidence for his 2003 invasion of Iraq was socially constructed.
For those of us who are not attached to a religion, we should assume that what we count as evidence still has a component that is socially constructed.
Personally, I hold that what I count as scientific evidence is not entirely socially constructed. But it is difficult to avoid the conclusion that there is a socially constructed component.
I think language is extremely important.
I just don’t think human language lends itself very well to formal logic.
At least not to things that are not or cannot be operationally defined.
Not just difficult, but impossible. Science itself is a social construct. The line between science and pseudoscience is very thin. Those who think they can demand evidence without defining how they mean it are socially incompetent, and with great likelihood also scientifically challenged.
Is it brains or minds? If it’s minds, then okay. If brains, then Andy Clark should know that a brain is a bunch of nerve cells and whatever it generates is a nerve signal and those can be measured and detected. Has he defined a type of nerve signal that he can call “prediction”? If not, he is not arguing. Not in any scientifically relevant sense anyway.
What is the “bottom-up direction” by which information is propagated? That information is coming from somewhere, right? Why is not possible to call it “sensory data”? What’s the difference?
The scheme you present looks fundamentally the same as any mind-body relation described since antiquity, just with some modern scientific terminology randomly added here and there, and some crucial fundamental concepts avoided equally randomly.
Your scheme is different if you think it’s material all the way or that there is no hierarchy or that there should be no hierarchy. There’s no other difference.
As rational people know, there is a hierarchy – and must be. “Monkey see, monkey eat” does not describe a human. A human would reflect whether to eat and what to eat, then search for what has been determined, and eat only then, not before.
I don’t depend on my worldview to justify my knowledge of science. I depend on God to justify all my knowledge.
He should not do that.
On the other hand since God has revealed himself there is no need to create a fiction to fill his role.
peace
The problem with that proposal is that you need to explain why you can be confident that your senses and cognitive facilities are up to the task of giving you accurate sensory data.
peace
What he needs to do is provide justification for his beliefs.
we all need to do that
So far he has been unable to do so.
By what authority do you make this pronouncement?
Are you acting as the decider in chief again?
What is it about you that makes you feel that you have this authority?
peace
peace
According to you.
So do you. All you have is assertion. God reveals it you say. You can’t prove that, it’s just something you believe. How do you know that god reveals it? God reveals that too? Que infinite regression of revelations.
You claim you know it, but all that really means is that you believe strongly. You can’t even demonstrate the truth of it to yourself.
This act of superiority you manifest appears as nothing more than that: an act. When asked to justify your assertions, you just pile more on top.
You obviously haven’t met any number of my relatives.
No he doesn’t.
When he reads something in Reason, he figures it must be true and, so, believes it. As he would only believe things with objective empirical evidence he figures he must have that as well. He thus infers that being reported in Reason is objective empirical evidence. Seems silly, I know, but keep in mind that you do the same thing with the Bible.
I think the top-top assertion needs to be detailed by relating it to how we learn and understand language. Written or spoken language is perceived and so processed by PP two-way mechanism as well, not simply top-down.
On language usage and development: Cognitive linguistics studies language understanding as an embodied process, even going so far as mapping metaphors to embodiment. It studies language learning as using the same mechanisms as are used for other types of learning. In both cases, PP processing would seem to apply. This is not purely top-top processing.
Embodiment and common learning mechanisms speak to continuity with animals.
I do agree that language gives a richer set of conceptual schemes which can be used as the top level models in two way PP processing, but I am uncomfortable with the restriction to top-top processing, and with the sharp boundary that some of your posts draw between all animals and humans.
Massimo Pigliucci’s latest book is hot off the virtual presses:
The Nature of Philosophy.
Subtitle: How Philosophy Makes Progress and Why it Matters
Because my cognitive factulties and my senses are tied together to aid my survival. If they were consistently inaccurate they wouldn’t be aiding my survival.
Done.
But petrushka has already decided this matter, as he’s repeatedly indicated. Neil too. They ‘HOLD’ this and that thesis, so books on the subject (especially by philosophers!) are not terribly relevant.
Note that KN’s approach is somewhat different: he tells us rather what he’d LIKE to hold (e.g. A theory of ‘knowing that’ which is not classist or otherwise hard on menial labor, an acquisition of language theory that utilizes both affordances and Nancy Cartwright, etc.). It’s a bucket list.
I’ve been following Massimo’s presentation on his blog posts.
Consider me unpersuaded. Philosophy evolves, in the sense of changing. But it does not make progress as far as I can see. It’s methods are largely the same as the methods of theology — making up stuff as you go along.
Neil Rickert,
There you have it!
Tell me how the non-verbal animal does this.
Also, how would a species of animal evolve so that its cognitive faculties would not be up to the task of giving accurate sensory data?
Beyond that, what if something goes wrong, like schizophrenia? How would such an individual find out that things had gone wrong? I’d note that many of them have no clue that they’re wrong. But you may have revelation that shows that they’re wrong, only they have revelation that tells them otherwise.
Glen Davidson
I still don’t agree with your first clause. In formal systems like mathematics, the rules for allowed logics could, I suppose, be considered to be social practices, broadly defined. Empirical evidence, though, is independent of those practices. That’s one reason why scientists from widely different cultures are able to reproduce each others results. At the level of non-human animals, it’s why those with mental models more aligned with reality survive.
I’m not sure you have the arrow of causality pointing in the right direction. I suggest that conceptual frameworks that are better aligned with reality provide a better basis for integrating empirical knowledge. Social norms can, at best, stay out of the way of objective evidence.