Joynewyeary

Reality, Causality, and Logic

36 posts in this topic

Suppose that some researchers want to confirm that some specified kind of event causes some other specified kind of event. It is possible that detailed knowledge of at least one specialized science will be involved in trying to establish such a thing. However, there is a general philosophical issue that arises. Are the researchers confirming merely that some specific observations are consistent with a particular model of reality? Is causality a definite aspect of reality itself?

There is a distinction between correlation and causation. How can one guarantee that one has discovered an example of causation and not merely a very subtle or complex correlation? Does the specialized science known as "probability and statistics" play a role in establishing causation or is there a more general logic of inductive inference that the science of probability and statistics relies upon, but that does not itself rely upon the science of probability and statistics?

Share this post


Link to post
Share on other sites
Suppose that some researchers want to confirm that some specified kind of event causes some other specified kind of event.

This is an incorrect Humean view of causality. Events don't cause events. Entities cause actions or, more precisely, entities act in accordance with their natures.

It is possible that detailed knowledge of at least one specialized science will be involved in trying to establish such a thing.

Maybe. The important thing is to understand the nature of the entity and how what something is determines what it does. Sometimes that understanding comes from simple observation and sometimes it requires advanced scientific research.

However, there is a general philosophical issue that arises. Are the researchers confirming merely that some specific observations are consistent with a particular model of reality? Is causality a definite aspect of reality itself?

Sure. Causality is the law of identity applied to actions. What something is determines what it does.

There is a distinction between correlation and causation. How can one guarantee that one has discovered an example of causation and not merely a very subtle or complex correlation? Does the specialized science known as "probability and statistics" play a role in establishing causation or is there a more general logic of inductive inference that the science of probability and statistics relies upon, but that does not itself rely upon the science of probability and statistics?

This is an area of philosophy yet to be developed. My own view is that you have identified the cause when you have found what it is about an entity that reduces a statement about its actions to a tautology. For more, see my postings on the Induction thread.

Share this post


Link to post
Share on other sites
entities act in accordance with their natures.

I vaguely recall reading somewhere that sleeping pills cause drowsiness because they contain at least one substance that has soporific qualities. I suppose that such an explanation may help expand the vocabulary of a student of English as a foreign language. However, I must admit that such an explanation provides no meaningful information that I can understand.

Share this post


Link to post
Share on other sites
Suppose that some researchers want to confirm that some specified kind of event causes some other specified kind of event.

This is an incorrect Humean view of causality. Events don't cause events. Entities cause actions or, more precisely, entities act in accordance with their natures.

If I assume that you are right when you say, "Events don't cause events [...] entities act in accordance with their natures" then should I conclude that people investigating an airplane crash should, for example, pay no attention to the question of whether or not the airplane was struck by lightning?

Share this post


Link to post
Share on other sites
Suppose that some researchers want to confirm that some specified kind of event causes some other specified kind of event.

This is an incorrect Humean view of causality. Events don't cause events. Entities cause actions or, more precisely, entities act in accordance with their natures.

Can you demonstrate that the view that events cause events is a view that is contrary to reality and reason?

Share this post


Link to post
Share on other sites
entities act in accordance with their natures.

I vaguely recall reading somewhere that sleeping pills cause drowsiness because they contain at least one substance that has soporific qualities. I suppose that such an explanation may help expand the vocabulary of a student of English as a foreign language. However, I must admit that such an explanation provides no meaningful information that I can understand.

It doesn't. A valid causal explanation might be in terms of the chemical properties of the sleeping pill and how it interacts with the chemistry of the human nervous system such that it results in drowsiness. The relevant "natures" here are what is in the pill and what the human nervous system is.

Share this post


Link to post
Share on other sites
Suppose that some researchers want to confirm that some specified kind of event causes some other specified kind of event.

This is an incorrect Humean view of causality. Events don't cause events. Entities cause actions or, more precisely, entities act in accordance with their natures.

If I assume that you are right when you say, "Events don't cause events [...] entities act in accordance with their natures" then should I conclude that people investigating an airplane crash should, for example, pay no attention to the question of whether or not the airplane was struck by lightning?

They shouldn't do so unless they have a reason to do so and to find such a reason, they have to look at the plane that crashed and the conditions around the plane when it happened. If it was a fair and sunny day and there are no markings or other evidence on the plane to indicate a possible lightning strike, then there is no reason to consider that a possibility.

When investigating the possible cause of an event, you look at the nature of the entities that acted.

Share this post


Link to post
Share on other sites
Suppose that some researchers want to confirm that some specified kind of event causes some other specified kind of event.

This is an incorrect Humean view of causality. Events don't cause events. Entities cause actions or, more precisely, entities act in accordance with their natures.

Can you demonstrate that the view that events cause events is a view that is contrary to reality and reason?

No, I can't prove a negative. Observe, however, that the primary advocate of the view that events cause events, David Hume, couldn't prove his view was correct. He had the burden of proof and failed to make his case.

Share this post


Link to post
Share on other sites
If I assume that you are right when you say, "Events don't cause events [...] entities act in accordance with their natures" then should I conclude that people investigating an airplane crash should, for example, pay no attention to the question of whether or not the airplane was struck by lightning?

I infer that you're implying that lightning is "an event" to disagree with the statement that "events don't cause events".

An event can be said to be some *thing* that is *acting at some particular time*. That actually mixes up entities with their actions. A more precise way to state "Events don't cause events" would be "Actions do not cause actions".

Lightning is a massive surge of electrical current, carried by electrons accelerated to a high voltage. In a philosophic sense, I think that a "bolt of lightning" counts as an entity. Then, in terms of *action*, it can cause thunder (caused by superheated air from the discharge creating a shock wave) and damage other entities such as metallic airplanes.

Lightning (entity) -> Action (energy transfer), resulting in certain effects on other nearby entities (airplane: electrical circuits, material damaged, etc.)

You could choose to focus on lightning's constituent particles - accelerated electrons - instead, but that doesn't really change the analysis.

Share this post


Link to post
Share on other sites
Suppose that some researchers want to confirm that some specified kind of event causes some other specified kind of event.

This is an incorrect Humean view of causality. Events don't cause events. Entities cause actions or, more precisely, entities act in accordance with their natures.

But that is in fact saying the same thing with different words, only less efficiently. Every time we observe a cause and effect in the world there is some physical interaction at work. Suppose I'm sleeping and I awake while a dog is barking somewhere. The event "barking" is the cause of my awakening. Now I could also say that the entity "dog" is the cause, as it is "in accordance with its nature" that it barks, but in fact this is less informative, as the dog may be usually quiet and it's only barking while there is a burglar prowling around. Is now the entity "burglar" the cause of the barking of the dog? Again this is less informative than saying that it is the event "burglar prowling around the house" that is the cause of the barking, you might say that it is in the nature of the burglar to prowl around that house, but that sounds rather contrived to me. He could as well be at home or burgling a house in a different quarter of the town.

Joynewyeary's example of the lightning striking a plane is even more telling. In his example it is not relevant how the investigators decide that it was a lightning strike that caused the crash, but the fact that it was a lightning strike. A lightning strike is a prime example of an event, a lightning is not an entity, the entities are electrons, ions, molecules, atom and photons in a very specific interaction. Saying that those particles are acting according to their nature isn't informative at all, there are zillions of other possible actions of the same entities, but it is only the highly specific interactions that constitute a lightning strike that is the cause of the plane crash.

No, I can't prove a negative. Observe, however, that the primary advocate of the view that events cause events, David Hume, couldn't prove his view was correct. He had the burden of proof and failed to make his case.

I don't know Hume's argument, but the fact that you can't prove his view (like a mathematical theorem) doesn't imply that it is incorrect. I gave just two (in fact three) examples of an event causing an event, and it's quite easy to extend it to millions of examples (given enough time and stamina). So he might arrive at his hypothesis by induction. If he's wrong, it shouldn't be difficult to give a counterexample.

Share this post


Link to post
Share on other sites

Please define 'event' if you hold that it can cause other events.

Share this post


Link to post
Share on other sites
Please define 'event' if you hold that it can cause other events.

A very specific interaction between entities. See my example of a lighting strike (interaction of the particles I mentioned). Or the barking of the dog (a complex interaction of physiological entities and the surrounding air molecules).

Share this post


Link to post
Share on other sites
No events occur without entities. Events refer to interactions of entities.

Sure, entities are an essential part of an event. But a cause A of an event B is never a cause without some kind of interaction and it is much more useful to describe that cause by its specific interactions than by merely referring to the "nature" of its entities, which is just a shortcut for all the possible interactions they can have, which in general will be a huge number. Only a very tiny fraction of all those possible interactions will lead to the described event B. If someone throws a stone, breaking a window, you might say that the stone is the cause of breaking the window as it is the nature of the stone that it can break windows. But it's also in the nature of a stone to act in millions of other ways without breaking a window and it will seldom fly on its own through a window, therefore it's much more useful to describe the cause as the event of someone throwing a stone, it is that very specific interaction of the stone that causes that event B, as it gives us much more relevant information and a better understanding of why event B happens. That person could for example also have thrown a book through the window, which illustrates that in fact throwing certain kinds of objects is the cause of breaking the window rather than the nature of a book. That doesn't mean that its nature is irrelevant, it must have certain characteristics to be able to break a window, a matchbox won't work for example, but saying that the book by its nature is the cause of the window breaking gives us very little useful information compared to a description of the cause as an event.

Share this post


Link to post
Share on other sites
A lightning strike is a prime example of an event, a lightning is not an entity, the entities are electrons, ions, molecules, atom and photons in a very specific interaction.

"My body is a prime example of an event, a body is not an entity, the entities are electrons, ions, molecules, atoms and photons in a very specific interaction."

Er - *any* macroscopic entity is necessarily comprised of what are generally a very large number of more fundamental particles. That doesn't mean that the entity as a whole doesn't have some identity or characteristics apart from other macroscopic entities.

Share this post


Link to post
Share on other sites
Please define 'event' if you hold that it can cause other events.

A very specific interaction between entities. See my example of a lighting strike (interaction of the particles I mentioned). Or the barking of the dog (a complex interaction of physiological entities and the surrounding air molecules).

If an event is "a very specific interaction between entities," then your definition already includes that entities are the primary actors in the interaction. All Betsy is saying is that the interaction is a result of the nature of the entities, and there would be no event without it.

So it does not mean that information is lost when one says it is the entities that caused the event. To hold that the event "barking caused awakening" is explanatory seems a lot less informative than "the barking dog woke me from my sleep." Why the dog was barking is irrelevant to the the statement. While lightning is not an entity, it is caused by the realignment of atomic material under certain physical conditions that have the nature of interacting with other physical objects.

Share this post


Link to post
Share on other sites
--------------

A lightning strike is a prime example of an event, a lightning is not an entity, the entities are electrons, ions, molecules, atom and photons in a very specific interaction. Saying that those particles are acting according to their nature isn't informative at all, there are zillions of other possible actions of the same entities, but it is only the highly specific interactions that constitute a lightning strike that is the cause of the plane crash.

------------

If "saying that those particles are acting according to their nature isn't informative at all," then by what jump do you assert that it is the nature of interactions to produce causes? Such a claim is less informative (if at all), for it is the entities that one is investigating when one observes interactions.

Share this post


Link to post
Share on other sites
No events occur without entities. Events refer to interactions of entities.

Sure, entities are an essential part of an event. But a cause A of an event B is never a cause without some kind of interaction

No one has put forth a statement disagreeing with that.

and it is much more useful to describe that cause by its specific interactions than by merely referring to the "nature" of its entities, which is just a shortcut for all the possible interactions they can have, which in general will be a huge number. Only a very tiny fraction of all those possible interactions will lead to the described event B.

I'm not sure how you come to this conclusion. While it is true that an entity has a large number of potentialities, which specific one was actualized under the specific conditions requires investigation and identification. If a plane crashes (one of its many potentialities) whether it was struck by lightning has to be ascertained by investigating the entities: one looks for evidence on the plane not by looking for a lightning event.

If someone throws a stone, breaking a window, you might say that the stone is the cause of breaking the window as it is the nature of the stone that it can break windows. But it's also in the nature of a stone to act in millions of other ways without breaking a window and it will seldom fly on its own through a window, therefore it's much more useful to describe the cause as the event of someone throwing a stone, it is that very specific interaction of the stone that causes that event B, as it gives us much more relevant information and a better understanding of why event B happens. That person could for example also have thrown a book through the window, which illustrates that in fact throwing certain kinds of objects is the cause of breaking the window rather than the nature of a book. That doesn't mean that its nature is irrelevant, it must have certain characteristics to be able to break a window, a matchbox won't work for example, but saying that the book by its nature is the cause of the window breaking gives us very little useful information compared to a description of the cause as an event.

This your argument is context dropping and/or a straw man. No one has claimed that stones or books spontaneously break windows. Of course they have to have some interaction. But "throwing" is no more explanatory than what you argue against. Throwing a rock doesn't break a window by itself and is no explanation. A thrown rock or book must actually interact with the window. An entity acts in accordance with its nature: when a rock is thrown, it can break a window when it hits it. Whether the rock was thrown by someone or by a tornado is irrelevant to the basic cause of the broken window.

Share this post


Link to post
Share on other sites
Suppose that some researchers want to confirm that some specified kind of event causes some other specified kind of event.

This is an incorrect Humean view of causality. Events don't cause events. Entities cause actions or, more precisely, entities act in accordance with their natures.

But that is in fact saying the same thing with different words, only less efficiently. Every time we observe a cause and effect in the world there is some physical interaction at work.

Not always. Sometimes there is just one entity and its actions.

If a ball rolls, it is rolling because it is round. It is acting, not interacting. Yes, a plane may crash if it is struck by lightning, but it can also crash if it runs out of fuel. Whenever you have an action, it is caused by an entity that acts in accordance with its nature. You may also have more than one entity and the entities may interact according to their natures, but that is not always the case.

Share this post


Link to post
Share on other sites
But a cause A of an event B is never a cause without some kind of interaction and it is much more useful to describe that cause by its specific interactions than by merely referring to the "nature" of its entities, which is just a shortcut for all the possible interactions they can have, which in general will be a huge number.

A ball rolls because it is what it is. The rolling is caused by the shape of the ball. The ball rolls because it is round. This is a simple and valid causal explanation and there is no "huge number of interactions" involved.

This kind of analysis applies to interactions as well. For example:

If someone throws a stone, breaking a window, you might say that the stone is the cause of breaking the window as it is the nature of the stone that it can break windows.

It is also the nature of a (glass) window that it is brittle and when it is struck with a certain amount of force, it will shatter. A cloth curtain, when struck by the same stone with the same force, would not break. To understand interactions you look at the nature -- i.e., the characteristics -- of the entities that interact.

Share this post


Link to post
Share on other sites

Ah, I now see where the problem lies. We are talking about "cause" in two different meanings. When I'm talking about causality, I use the common meaning of an event A causing event B where B happens at a later time than A, where we can distinguish a sufficient cause and a necessary cause. We say for example in special relativity that A cannot have a causal effect on B if A lies outside the light cone of B (whatever the influence is, it cannot travel faster than the speed of light). I suppose this is also the meaning that Hume used. However, you use "cause" in the sense of a scientific model that explains a certain phenomenon. Take for example the rolling ball: the explanation that a spherical object rolls easily and a cube not is that due to the geometry of the sphere the movement of the center of mass is in a horizontal plane so the ball needs no energy to counteract the gravitational force and as the contact surface between ball and support is very small, there is little friction. A cube however has a large contact surface, so it can only slide if the surfaces are very smooth (as on ice for example), and for rolling the cube, it must raise periodically its center of mass and that demands an extra amount of energy, therefore a thrown die will not roll as long as a ball. You might summarize that explanation as "the cause of the fact that a ball rolls is that it is round". But that is simply a different meaning of "cause" than Hume uses; an example of his use of "causality" would be: the cause that ball rolls is the fact that someone hit it with a cue (first the hit then the rolling, if he doesn't hit it, it won't roll). So you can't say that Hume's use of the term "causality" is incorrect, he simply uses a different definition, namely as a relation between two specific events, not as a general (scientific) explanation of a certain phenomenon.

Share this post


Link to post
Share on other sites
Ah, I now see where the problem lies. We are talking about "cause" in two different meanings. When I'm talking about causality, I use the common meaning of an event A causing event B where B happens at a later time than A, where we can distinguish a sufficient cause and a necessary cause.

Actually, when you use it in that (Humean) sense, there is no way to determine that anything is a sufficient or necessary cause.

Hume begins with a single instance of causality. All we observe is one billard bill hitting another and the second moving. "This is the whole that appears to the outward senses. The mind feels no sentiment or inward impression from this succession of objects: Consequently, there is not, in any single, particular instance of cause and effect, anything which can suggest the idea of power or necessary connexion."

http://oregonstate.edu/instruct/phl302/dis...e/comment2.html

However, you use "cause" in the sense of a scientific model that explains a certain phenomenon. Take for example the rolling ball: ...

Causality, as the relationship between an entity and its actions, is what explains all actions and interactions. What other sense of the term is there?

But that is simply a different meaning of "cause" than Hume uses; an example of his use of "causality" would be: the cause that ball rolls is the fact that someone hit it with a cue (first the hit then the rolling, if he doesn't hit it, it won't roll). So you can't say that Hume's use of the term "causality" is incorrect, he simply uses a different definition, namely as a relation between two specific events, not as a general (scientific) explanation of a certain phenomenon.

Hume's "relation between two events" is -- according to Hume -- unperceivable and not necessary. What, then, is Hume's "causality" good for?

Share this post


Link to post
Share on other sites
Ah, I now see where the problem lies. We are talking about "cause" in two different meanings. When I'm talking about causality, I use the common meaning of an event A causing event B where B happens at a later time than A, where we can distinguish a sufficient cause and a necessary cause.

Actually, when you use it in that (Humean) sense, there is no way to determine that anything is a sufficient or necessary cause.

You cannot logically deduce that A is a sufficient or necessary cause of B, but we can create a model with the hypothesis that A is a cause of B, and we can test that model and the hypothesis - that is what we do in science. Example: in 1820 Ørsted discovered that switching on an electrical current (event A) would deflect a compass needle near the electrical wire (event B). This turned out to be a systematic effect, so the plausible hypothesis was that event A is a sufficient cause for event B. Ørsted's model went no further than a description of the effect. Faraday realized its importance and did further research and created a more extensive theory which was finally perfected by Maxwell (which in fact confirmed that A is indeed a sufficient cause for B). Ørsted's discovery of a causal relation between events A and B was of course only a first step, but first steps are also important, even if they explain little in comparison to later theories that are built on that first step. Many scientific discoveries started with the discovery of a causal relation between certain events and much of our intuition that we acquire in daily life is based on the same principle (if I do this, invariably that will happen), which enables us to deal with the world around us, even without knowing all the scientific details of what happens, which would be impossible if we couldn't infer causal relationships between events.

Causality, as the relationship between an entity and its actions, is what explains all actions and interactions. What other sense of the term is there?

In the sense in which you use it, it is a synonym for "scientific theory". But in science and in daily life it is used in a different sense. See for example here, here and here.

Share this post


Link to post
Share on other sites
Ah, I now see where the problem lies. We are talking about "cause" in two different meanings. When I'm talking about causality, I use the common meaning of an event A causing event B where B happens at a later time than A, where we can distinguish a sufficient cause and a necessary cause.

Actually, when you use it in that (Humean) sense, there is no way to determine that anything is a sufficient or necessary cause.

You cannot logically deduce that A is a sufficient or necessary cause of B, but we can create a model with the hypothesis that A is a cause of B, and we can test that model and the hypothesis - that is what we do in science. Example: in 1820 Ørsted discovered that switching on an electrical current (event A) would deflect a compass needle near the electrical wire (event :). This turned out to be a systematic effect, so the plausible hypothesis was that event A is a sufficient cause for event B. Ørsted's model went no further than a description of the effect.

Certainly an hypothesis is usually the first step in discovering a causal relationship, by it is not itself a true causal explanation.

Faraday realized its importance and did further research and created a more extensive theory which was finally perfected by Maxwell (which in fact confirmed that A is indeed a sufficient cause for :).

How did he do that? Wasn't it by investigating the nature of the entities which acted and interacted and discovering which of their properties gave rise to their actions and how?

Ørsted's discovery of a causal relation between events A and B was of course only a first step, but first steps are also important, even if they explain little in comparison to later theories that are built on that first step. Many scientific discoveries started with the discovery of a causal relation between certain events and much of our intuition that we acquire in daily life is based on the same principle (if I do this, invariably that will happen), which enables us to deal with the world around us, even without knowing all the scientific details of what happens, which would be impossible if we couldn't infer causal relationships between events.

There is a difference between inferring a possible cause of an action and identifying what that cause actually is. In order to do the latter, you must understand the entities involved.

Causality, as the relationship between an entity and its actions, is what explains all actions and interactions. What other sense of the term is there?

In the sense in which you use it, it is a synonym for "scientific theory". But in science and in daily life it is used in a different sense. See for example here, here and here.

I realize that others use the term "cause" to stand for something different than "cause" in the Aristotelian / Objectivist sense, but omitting or not investigating entities never gives you a real causal understanding of things and events. It is a cognitive dead end.

Share this post


Link to post
Share on other sites
If a plane crashes (one of its many potentialities) whether it was struck by lightning has to be ascertained by investigating the entities: one looks for evidence on the plane not by looking for a lightning event.

Are you assuming that the airplane under discussion is not equipped with devices that measure and record various kinds of information over a period of time? For example, isn't it possible that an experimental airplane so equipped might be deliberately flown by a pilot, with no passengers on board, into lightning storms to record what happens?

Share this post


Link to post
Share on other sites