For example consider Newton’s second law P = mf. Here P is the force applied on a body of mass m producing acceleration equal to f. My question is should there be a temporal ordering between the instants of the application of P and the production of acceleration f? If it exists then there must be a delay present between the two happenings. On the other hand if it does not exist then how the sense of dependency can be maintained? How to avoid such logical contradiction?
Krishnan is correct in the temporal asymmetry. Cause must precede effect. The length of time of the process varies with the process. Infection precedes disease, but incubation times vary. Mechanical or electrical processes are often perceived as instantaneous. Calculation of the effects can be considered as instantaneous without loss of generality or introduction of error. Nevertheless, there is a time lag before full effect is achieved. This time lag is often the basis for control.
A bouncing ball must have time for compression and decompression, yet we assume a perfect, instantaneous change of direction.
Dear friends/followers, according to General Relativity a body on Earth surface is accelerated upwards. However its velocity remains zero. An electric/ecological car, which you have left in the garage is being constantly accelerated (g = 9.8 m/sec^2) without being in motion (v = 0). To your case now. The moment you change the force P, the velocity remains the same. I am sorry for being too logical, perhaps. Bye!
The cause-effect relationship holds in many but not all cases. That is why we describe a well-established theorem, in this respect, as a law. In daily real life, we may do all the necessary things (as far as we know) & thereby we are applying the cause but the effect or the result may or may not occur. For example, you may prepare a land well and put the seeds but this does not ensure getting a good crop in all cases.
I mentioned Newton's Second law as an example. I think my statement was not sufficiently illustrative. I therefore should prefix "For example" to the explanation part.
I don't see how by definition the cause and the effect of the cause can occur at exactly the same time
When you connect a battery across an ideal resistance with no stray capacitance and/or inductance, current starts flowing through it instantly. At least we understand ideal resistance like that. Is it then we are not allowed to even conceive an ideal resistance.
I think in Social Sciences, the relationship is causal. Cause and effect is a relationship between two or more events or things, where one is the result of the other or others.
I would say that there is one process, not several processes. Not a process of a cause and later a process of its effect. Nietzsche made a joke saying something like “The cause pulls and pulls until the effect occurs.” It is of course not happening that way.
Maybe there is some friction hindering the acceleration, but that is not before the process in a temporal sense. It becomes relevant just in the moment the (main) force starts.
The notion of causal signal is needed in physics only on an operational construal of that; on a less operational view, notions like flow of energy-momentum and various temporal notions such as the light cone structure suffice for the purposes that talk of causal signals have been standardly put., but of course from the nonappearance of the notion in fundamental physics it doesn't follow that fundamental physics doesn't provide the means to explain it; and it certainly doesn't follow that the whole idea of causation is incompatible with what fundamental physics tells us.
The relation between cause and effect is supposed to have an important temporal asymmetry: causes normally or always precede their effects. This does not appear to be simply a matter of the earlier member of a cause-effect pair being conventionally called the cause; rather, it is connected with other temporal asymmetries that play an important role in our practices.
For instance, we tend to explain later events in terms of earlier ones but not vice versa; and we think that it makes sense to stop smoking as a teenager so that one will not get cancer later, but that it does not make sense to take a cancer-preventative later in life so that one will not have smoked as a teenager (or to take a cancer-preventative in childhood so that one won’t smoke later on). Most people would defend these practices on the grounds that causes explain their effects but not conversely, and that it makes sense to prevent an effect by preventing its cause but not vice versa. The notion of cause is intimately bound up with these asymmetries of explanation and action, as well as with numerous other temporal asymmetries.
http://philosophy.fas.nyu.edu/docs/IO/1158/Cause.pdf
Very nice question, congratulations!
It is supposed that this kind of questions follow the general guide that information is being transported with a velocity less than speed of light, according to relativity theories.
But, has anybody really measured the time between cause (force) and result (acceleration)?
I'd like to see such a data set please...
Krishnan is correct in the temporal asymmetry. Cause must precede effect. The length of time of the process varies with the process. Infection precedes disease, but incubation times vary. Mechanical or electrical processes are often perceived as instantaneous. Calculation of the effects can be considered as instantaneous without loss of generality or introduction of error. Nevertheless, there is a time lag before full effect is achieved. This time lag is often the basis for control.
A bouncing ball must have time for compression and decompression, yet we assume a perfect, instantaneous change of direction.
Also between that cause and force time can be hidden present as well as the mind of the observer.
Let be quote from Stanford encyclopaedia:
“Sometimes also called retro-causation. A common feature of our world seems to be that in all cases of causation, the cause and the effect are placed in time so that the cause precedes its effect temporally. Our normal understanding of causation assumes this feature to such a degree that we intuitively have great difficulty imagining things differently. The notion of backward causation, however, stands for the idea that the temporal order of cause and effect is a mere contingent feature and that there may be cases where the cause is causally prior to its effect but where the temporal order of the cause and effect is reversed with respect to normal causation, i.e. there may be cases where the effect temporally, but not causally, precedes its cause.”
http://plato.stanford.edu/entries/causation-backwards/
Dear All,
The general or clear order of cause and effect often cannot be recognised. First because the delay is so long that the true cause cannot be identified. Another disturbing factor may be the complexity of causes and the observer’s imperfectness, both influence the observation. Regarding Marcel’s purely theoretical comment, a naive observer often is not able to recognise either the cause or the effect depending on the intensiveness of phenomena. It is very difficult to realise and understand the causative order or sequence. The recognition of causative order - regarding simple events and mechanisms - may be the “easiest” in physics and the most complicated in the social sciences. That is why the physics is the most developed scientific field.
Causal relations occur between two events. If complex events, being aggregations of smaller microscopic events are anlyzed, they may occur in the same time (as far as we may observe them).
Consider such statement: "In 2011 economy growth caused increased consumption of luxury goods". Both events (A: economy growth) and (B: are increased consumption of luxury goods) are macroscopic, quite vague to define and their start/end times are blur enough to state that they are simultaneous (in 2011). Of course, at the microscopic level casualty and temporal precedence is reather kept.
This can be mapped on ideal gas law, PV=nRT: pressure grow causes that temperature raises (simultaneously).
Blending Remi's and Piotr's comments I would say that there is a difference between 'dumb' particles/masses in physical processes and intelligent actors in social processes. In contrast to particles the intelligent actors may anticipate possible or likely effects of their actions, which causes them to react before the cause actually happens.
Why do we talk about reducing CO2 and other emissions? It is because we want to prevent the potential effect that, otherwise, by nature would lead to a new stability in the atmosphere. Given we can prevent global warming, this would be caused by the possible effect and the actions taken. Is there something like that in the world of particles?
Regularity analysis can succeed in distinguishing genuine causes from effects, epiphenomena, and preempted potential causes—and whether it can succeed without falling victim to worse problems, without piling on the epicycles, and without departing from the fundamental idea that causation is instantiation of regularities. Dynamical laws that appear to govern our world, tells us that there would be some extremely tiny (but non-zero) probability that the ice cube would grow, while the water heated up. Time-reversibility tells us that such an occurrence is nomologically possible, statistical mechanics that it is extremely unlikely.
http://lapaul.org/papers/Introduction.pdf
Anup spoke of a dependency. A (necessary) dependency can be expressed by
If not A, then not B.
This is not a temporal relation, of course, because nothing is happening. But “If A then B” does not have to be temporal either. It can be about the hair color of twins.
Logical necessity:
It is necessary, that a cause has an effect. When A has no effect, it cannot be the cause. But that is no question of physics. It is sometimes called an “analytical truth” that the cause has its effect, i.e. true because the words have these meanings. That means it can turn out that A was not the cause, but it cannot turn out that the cause was not the cause. It is only a question of words. Therefore I think logic is not decisive in this discussion.
Thank you all for such elaborate discussion. My concern is to know how the dependency relation between the two events could be expressed. Let a system be defined by its initial state y0. Let the final state be yF that we will get when an input x is applied to the system. The final state is determined from the initial state and the input. After the input is applied state transition will start and the system will finally arrive at state yF. This may take a finite amount of time and people may argue that this delay will account for the necessary dependency relation. However there are systems where the beginning of the state transition will start from the instant the input is applied. If we consider this beginning as the output(effect) corresponding to the applied input (cause) then the problem of simultaneous occurrence can not be ruled out. Should the dependency does not necessarily requires temporal ordering?
Consider the popular toy that comprises a row of 10 steel balls suspended by strings. The balls are in contact and in a straight line. Lift one ball at the end of the row and let it fall against the other 9. The ball on the opposite end pops off at the same velocity as the first made contact (ideally). Was the transfer of momentum instantaneous? It was fast, but not instantaneous. Remove one ball and repeat. The momentum transfer is faster. Continue removing balls and repeating until there are only two. When there are two, is the momentum transfer instantaneous? Fast, but not instantaneous. The momentum must transfer from the center of mass of one ball to the center of mass of the other ball. The rate of transfer depends upon the material characteristics of the balls.
The transfer of momentum begins at the instant of contact of one atom each of each ball. We can see this by reducing the size of the balls until each is one atom in size. We still have to transfer the momentum from the center of mass of one atom to the center of mass of the other. Effect still follows cause.
Since we are imagining things, let us shrink the atom. We can shrink the atom until it is an infinitesimal point with no mass. Now cause and effect are simultaneous, but there is no effect or cause.
Should there exist a temporal ordering between a cause and its effect?
Yes, if you really mean cause-effect relationships
But Newton’s second law is NOT a cause-effect relationship.
I learnt Newton's 2nd law as a bidirectional, i.e. there is no cause and no effect, but perfect correspondence:
There is acceleration if and only if a force acts on the system, so no first and second.
Now this law only holds for perfect rigid solid bodies (likely in existing in nature...)
so in reality a succession of processes occur until the effect can be observed.
E.g. applying the force, compressing the object, increasing the internal force, reaction, supra threshold reaction beyond the static friction, movement...
Should there be a temporal ordering between the instants of the application of P and the production of acceleration f?
Again it depends on the exact particularities of the system under consideration, but in an ideal system this t is zero.
In my naive view, there is no logical contradiction
But if you really refer to cause-effect, then naturally cause must precede effect.
The length of time of the process varies with the process.
The Newtonian 3-body problem as already pointed out, is an excellent example of how limited our notion of cause and effect is.
Newton's first law concerns inertia, the second force and acceleration, the third action and reaction. All three laws are cause-effect relations. It is not obvious in inertia, because there is no mention of cause, only an effect. Nevertheless, a cause exists. The second law concerning force and acceleration is a definite cause-effect. Even if the force is insufficient to cause acceleration of the whole mass, it causes deformation and shock waves. The third law involving action and reaction is definitely a cause-effect law. The placing of a weight on a table is an example of action and reaction, two opposing forces without perceptual movement. The weight causes stresses and strains in the table, hence movement. Cause and effect.
The 3-body problem does not limit our notion of cause and effect, it limits our ability to calculate the movement exactly. Yet, we can calculate the positions of the bodies probabilistically to high precision.
Since the introduction of nonlocality cause and effect has lost its meaning.
The terms initial state, final state and input has in nature no sense, it is always about a multiplicity of embedded factors in time or out of time.
I think that to see scientific laws in terms of cause and effects is a wrong way to understand what they mean. Let us take two examples : the law of perfect gas PV=nRT. Could we say that temperature is the cause and pressure the effect ? Or the Kepler laws which describe the movement of planets. Actuallly the idea of causality is an heritage of Artistolecian philosophy. It is true that when one considers any physical system, with several objects whose properties are represented by variables, usually when one does an experiment some variables are known, and to check the law one computes the unknown variables, but by definition this can be done with any one of the variables, there is no reason to priviledge one vs the others.
Actually one of the most important principle of physics, the principle of least action, states that at equilibrium the value of the variables are linked all together, by the stationarity of some action. It underlines the fact that all the variables have to be considered on the same footing. And this is the only way to avoid all the disturbing questions whch arise in the relativist context. But this leads to consider systems, and physical laws, as encompassing the whole evolution of the system. Which means that "instatenous" laws are deduced from the results which are measured after the acion has occured. The planets have their trajectories because some equilibrium has occured between the llaws governing their interactions, and the pressure, temperature of the gas have a meaning because some equilibrium has been established. Of course it is nice, and convenient, to use derivatives and deifferential equations, but they are just the mathematical consequences of the facts which are seen.
@jean claude Dutailly
I do not understand why Newton;s Second law can not be considered as a cause and effect relation. From the first law we find, When viewed in an inertial reference frame, an object either remains at rest or continues to move at a constant velocity, unless acted upon by an external force. The law states clearly that a change of state viz., "at rest"or "at a constant velocity" can be changed only by the application of an external force. The second law gives a measure of this change due to the applied external force. The law of perfect gas PV=nRT. may be considered as a perfect gas system invariant. However when you change one of the parameters the others also change keeping the invariant valid. Therefore the cause and effect relation is important when one studies the system dynamics, where a physical law depicts the invariant for the system. Invariant and the system dynamics are two different things. The first one defines the system characteristics where the second one describe the system behavior when executed. However during execution a system invariant should always remain valid.
The cause effect relation consists between events, not within the law. The laws do not provoke an effect. They rule a process in the sense that the process evolves as depicted by the laws. In this sense laws rule everything, but they do not stimulate anything, they do not evoke an event. Otherwise we would live in a totally determinist world.
Maybe vacuum fluctuation is an exception? They evoke something. But that might be called a principle, not a law. What would you say?
In answer to the question of respected jean claude Dutailly " the law of perfect gas PV=nRT. Could we say that temperature is the cause and pressure the effect ?", No we cannot say that. The equation of state for an ideal gas describes the relationship between 4 variables (pressure, volume, number of moles, and temperature). No variable can have a control on the gaseous state to be described as a cause.
Yes the law states the relationship among 4 different variables and thus it does not describe a cause and effect relation. But when some of the parameters (not all) are varied the values of the other variable will change. This is execution of the given dynamic system and here comes the cause and effect relationship.
Of course in alaw such that PV=nRT when one changes a parameter one can consider that the change on one has an effect on the other. But this is misleading. As for many laws it is valid in a state of equilibrium, so actually what one can see is the move from one equilibrium to another. To represent this move through a differential is convenient, but does not correspond to the physical phenomenon, notably when the transformation occurs through inhomogeneous states.
If I have understood correctly what you say is that the law depicts the relation among different parameters at equilibrium. When some change of a parameter value is made (cause) a state transition will start. It may be during the transition one may not assume the relation that is valid in a state of equilibrium. I could not understand how such transition denies the existence of a cause and effect relationship, however complicated it could be.
In the very same relationship PV=nRT, doubling the number of moles of a gas by ,say injection, at constant T& P, indicates doubling the volume of the gas simultaneously. This is not cause-effect since 2 moles of a gas means 2 volumes of a gas.
Doubling the number of moles is an action and thus the volume in the system becomes double then why there should not exist a cause - effect relation? There is one ball in a box you add another and the count becomes two. Do you think there is no cause and effect relation.
Exact: In general, scientific laws are correspondences and not "cause and effects".
Even if we consider 2nd Law of Newton again:
F=m*a
and we read as cause-effect, then actually mass and acceleration would be causes of force, and not viceversa...
but
a=F/m
reads as cause-effect, force and acceleration would be causes of acceleration
but actually is correspondence
there is F if and only if there is m and a
usually physics are not able to determine root causes but the describe the relationships
in this sense, I agree to Martin and Jean Claude.
"The cause effect relation consists between events, not within the law. The laws do not provoke an effect. They rule a process in the sense that the process evolves as depicted by the laws."
I would say laws need variables – experiments need values. No values in laws, only relations between variables. As far as I remember, there is only one constant (using a value) in laws, G. (I suppose c should be added as a constant that can occur in laws). Laws are always describing relations, therefore there cannot be a simple (primitive) symbol for a law.
But these variables are not “in the laws”, they are mentioned by them. Some authors call laws a relation between universals (e.g. D. Armstrong). Universals instead of objects. The objects cannot occur in the laws, naturally. A universal could be “mass” - not “this mass”. Therefore there is no causation among the elements that occur in a law statement.
Whatever the topic, a scientific law in Physics can be represented as follows :
you have a system, comprised of several objects (solid, fields,...) with their properties, whose values are expressed by variables X1,X2,..Xn. The law states that, whatever the occurence of the system, the initial cnditions,...we have a given relation between the variables. Uusually in an experiment on keep one of the variable fixed (then it can be considered as the cause) and stydy how the other vary (then they are considered as the effect). But this is purely conventional.
Dear Martin,
I think according to your views in a law their exist no dynamics. It only tells us the relationship among different variables. Can we consider this law to be a system invariant? This relation will hold in equilibrium state. I admit. But in an experiment where system dynamics is involved some disturbance is necessary to start a state transition.
I would like to ask .Jean then why I should not be allowed to call this disturbance a cause.
Possibly then system dynamics related to the second law the force is not the cause rather application of the force should be considered as the cause.
Newton's laws are simplifications. They are models of physics. The IDEAL gas law is a model of a static situation. The laws are models of universally observed phenomena. Excellent comment, Martin.
The ideal gas law has a dynamic form. It must have to transition from one state to another. I once had an engineer tell me PV=nRT prevented control of gas flow in a pipe because any change must be instantaneous. Please note, a pulse change in a gas cannot travel faster than the speed of sound in the gas.
There is cause and there is effect. Models are models. They are not real, but some get close to describing reality in a well controlled situation.
Dear all,
I find it very difficult to doubt that the mathematical formalisms are a kind of idealization. The point goes back to the origin of mathematics. Do we find any actual triangles or squares or points or one-dimensional lines in nature? The answer is negative. Even if there were such things, our means of measurement would not allow us to distinguish them from close approximations--the margin of error would always be too large. Physical objects can surely be approximately square or triangular, and etc. They may be so little different from a square or triangle, at the scale that interests us, that it is safe to ignore the difference in making calculations. But this is surely not the same thing as there being perfect squares or triangles. Geometry apparently arose from practices of land measurement, required or useful after the start of agriculture. The mathematics basically saved on making measurements, since once particular measurements were made, other needed dimensions could be calculated.
In a somewhat similar way, physical theories, these days, are usually thought of as having reference to scales or measurement or energy. So, GR is a very good theory, adequate to many or most purposes --so long as we do not get too close to the Planck scale. But beyond a certain scale, you run into the problem presented by the "singularities" of GR--which theory predicts points of infinite density. This reflects the degree of mathematical idealization involved in the representation of the curvature of space-time. Its taken as a signal, to the effect, that no one really knows what is going on under conditions of extreme density. It invites further theory.
In consequence of these reflections, I tend to think that where the idealizations of physical theory, as in Newtonian mechanics, invite us to ignore the element of temporal sequence in the usual concept of causality, then we are entitled to be a bit skeptical. A stronger degree of conceptual conservatism recommends itself.
H.G. Callaway
To Martin although clearly out of the direct scope of this thread:
There is a possible/probable future revision of the SI...
"At its 24th meeting (October 2011) the CGPM adopted a Resolution on the possible future revision of the International System of Units (the SI). This Resolution takes note of the CIPM's intention to propose a revision of the SI, and sets out a detailed road-map towards the future changes.
Resolution 1 of the CGPM (2011): On the possible future revision of the International System of Units, the SI
In the "New SI" four of the SI base units, namely the kilogram, the ampere, the kelvin and the mole, will be redefined in terms of invariants of nature; the new definitions will be based on fixed numerical values of the Planck constant (h), the elementary charge (e), the Boltzmann constant (k), and the Avogadro constant (NA), respectively. Further, the definitions of all seven base units of the SI will also be uniformly expressed using the explicit-constant formulation, and specific mises en pratique will be drawn up to explain the realization of the definitions of each of the base units in a practical way.
While remarkable progress has been made over the last few years, the conditions for adopting the redefinitions, as set by the CGPM at its 23rd meeting (2007), have not yet been fully met. The CGPM encourages National Metrology Institutes, the BIPM and academic institutions to maintain their efforts towards the experimental determination of the fundamental constants h, e, k and NA."
http://www.bipm.org/en/measurement-units/new-si/
Anup,
If I understand “system invariant” correctly it is a property of a system (in relation to time or shift). If this is true, then it is not part of the laws and not a law in itself. I would say properties can sometimes be described by laws. This happens after many experiences (and experiments) with the objects.
The properties are not fixed by the laws - they are fixed by the world, if we like to say so. The laws can describe how the properties are fixed if something is the case. Some people say, laws are not even an explanation, they only tell what is going to happen next, when we do something in a well-defined frame (maybe a well prepared experiment). The frame is not part of the law either.
More important: the properties of the objects do not occur in the laws. Sometimes we can learn about the properties after we learned from the laws, how the objects behave (provided nothing interferes).
It is not a law that gold has atomic number 79, it is a description of a fact or simply a sentence. Facts and individuals do not occur in laws.
Samuel,
I will have to think about it, but it seems to me that these definitions using invariants are units needed in measurements. That would then be more “on the part of the experiments” than on the side of laws, is that correct?
@Martin, in my humble opinion absolutely.
I only wanted to point out that official physics do include some more "magic numbers" taken as they are.
But I fully agree to your view and explanation (also supported in my previous contributions to this thread). For me laws only represent relationships between entities (and further to that, but this is also another thread only hold true conventionally until they proof to be wrong)
It seems that causal relations occur between two types of entities: classes of events and their instances. Moreover, there are two types of such relations.
Let me give an example: consider the folowing statements:
1) Excessive speed is a primary cause of traffic fatalities.
2) According to police investigation, the fatal accident […indicated place and time...] was caused by the excessive car speed
The first (1) relates two classes of events: "driving with excessive speed" and "fatal traffic accident". The second (2) relates two instances of these events, which ocurred in a particular place and time, in which also a particular car and people were involved.
The first causalty relation (lets call it CC) is in fact a result of generalization of several observations of second type. It represents a belief about causes and effects. This relation is atemporal (classes/concepts are atemporal).
The second relation (CI) links two instances of events. As these events are attributed with time, e.g. start/end time, CI should be consistent with temporal relations between them. At least an effect should not precede the cause.
Knowledge about causes and effects is expressed using CC statements, eg. technical failures causes accidents, using a mobile while driving, causes accidents, etc. We develop such knowledge due to "mental habit" mentioned by Stefan Gruner.
Typically, to assess what was the cause of an event instance, we analyze known, most probable CC relations. E.g. an accident (an event instance) might have been caused by (a) excessive speed or (b) faulty breaks. However, if breaks were damaged during the accident, due to the temporal inconsistency the hypothesis (b) would be rejected.
Dear Martin,
I have borrowed the concept of system invariant from programming discipline that was introduced by E. W. Dijkstra. Let me explain what I meant by an invariant. A system may at any instant be defined by its state. Since state definition for a system may not be unique, I define this by the values of all the system variables at any instant. An invariant I is a predicate expression describing a relation among these variables that must be true if the system is found in equilibrium which is a legal state of the system. There may be more than one equilibrium (legal) state of a given system. I call this a legal state because if the system is pushed (at this moment let us not specify the reason) to some other state a stable system will be moved to one of its legal state after a finite amount of time. Therefore, The law of perfect gas PV=nRT. may be considered as a perfect gas system invariant.
A state transition will occur if some external input is applied. The final state of the system will depend on the external input and the initial state of the system. This transition can be modeled by a state transition rule.
I do admit that the whole thing is a model of the actual system, because we can not prove the exhaustiveness of the set of variables we considered and we do not know whether the Invariant definition is appropriate. But so long as the qualitative behavior of the system at least can be predicted by this model, we can accept that.
Forces felt as a body accelerates and decelerates can be described in multiples of gravity, or G.
G-force is simply a descriptive measure of acceleration. When stationary, the force felt by Earth’s gravity is 1G, however when a body undergoes a change in speed and direction, that force increases in proportion to the rate of change. The magnitude of these forces involved can easily exceed the Earth’s gravitational force.
Airplane cockpits then were all designed to withstand 18G impacts because if the person was already dead, why invest in stronger materials and structural support.
http://csel.eng.ohio-state.edu/voshell/gforce.pdf
It seems that most of the comments stipulating that a cause must precede an effect are using notions of cause equivalent to an "efficient" cause, in Aristotle's terminology. However, if you think of causality in terms of a "final" cause or "telos", then the cause could come AFTER the effect. For example, electing a desired candidate causes a citizen to cast a vote.
Such a cause might be equivalent to a "reason", and perhaps people don't like to think of causality this way because it imputes "reasons" to nature, which some people don't like to do.
To Anup,
I agree with your scheme, and actually it can be extended beyond programming. See my paper on Common structures in Scientific theories on this site. You will see how one can deal with the issue of the set of possible states.
To all,
Actually the focus put on causality, seen as the most important feature of logic, and then in the building of science, is misguided. It leads, as some philosophers (such as Schopenhauer) to deny the possibility of science itself : if we need to look for causes to explain the effects, this quest is endless, and we must conclude at the impossibility of science, as an explanation of the real world.
There are two distinct steps in Science.
Scientific laws are build from concepts (such as particle, gas, solid,...) related to identified phenomena to which are attached properties, which are related to the measures (represented by variables) that can be done about the phenomena. A system is just a collection of objects, such identified,with their properties, and is desribed by a model. This is the natural setup of experiments. It is linked both to the abstract side (the concepts) and the real side (the measures) of any physical study. A scientific law is just a statement that, whatever the system, some relation occurs between the variables. It can be checked and the falsifiability of the statement in any experiment provides the scientific characteristic of the law.
But by itself a scientific law does not provide an explanation, seen notably as a chain of events which could be described in a causality predicate. To go further the physicist build theories, which start from fundamental concepts, use the power of mathematical formalism, to prove other laws from more fundamental laws (or assumptions). An example : the Kepler laws state just some charactristics of the orbits of planets. They do not provide an explanation. But by combining Newton's gravitation and mechanics, Galilean geometry, one can deduce the Kepler's laws. Of course the issue of "Why is it so ?" stays, but it stays at an upper level, which is supported by a larger batch of proven laws and experiments. And a theory can be challenged, and replaced by another one. However this occurs in some clearly identified conditions : relativity replaces galilean geometry. In this process causality does not have any specific role :actually the key tool is the rule of inference upon which are based all mathematical demonstrations.
But this shows also that cosmology, with the pretense to be a theory explaining the whole universe, and its evolution, is a dream : at best one can, through astronomical observations, find a plausible explanation for what we see, but it woill never be a scientific law : one cannot check it by experimenting with another universe !
Dear Dr. Gazi Islam, Your comment is a great one. Old philosophers, including those who belonged to our region, understood that there is a "must" in the cause- effect relationship, i.e. if the cause precedes then the effect is bound to follow. However, the cause "as we observe in reality" may be inefficient or it may occur but the effect does not necessarily happen or it may (as you said) occur after the effect or both may occur simultaneously "as the simultaneous change of V & n in the equation of ideal gas PV=nRT ". Therefore, the way out of this dilemma is to consider the matters case by case without generalization.
Dear jean,
Thank you for your detail answer. But I am unable to follow the last part, that is "In this process causality does not have any specific role :actually the key tool is the rule of inference upon which are based all mathematical demonstrations" . Can you please explain this in some more detail.
Regards,
Anup
Dear jean, there are other views about relativity and the 'replacement' of Galilean geometry, see the work of Johan Prins, here in RG, or look this review:
https://www.researchgate.net/publication/265471655
"The key-point of the book is criticality of “the reference frame in which the Physics is actually occurring” and the result that we can look there from outside and use the common Lorentz transformations only to see there and not vice versa. The relativity is only unidirectional, otherwise we end up to several inconsistencies and non real physical phenomena."
Many logical 'traps' can be overcoming by the above view.
As for the "...one cannot check it by experimenting with another universe !", why are you so sure about? Haven't you ever read Religion Texts? (not 'word-by-word', but with a critical mind?)
I think that the requirement of causality is crucial and cannot be by-passed so easy.
If you look that view you will find a possible reason for logical traps:
Data A five page review of the August 30, 2014 version of the Boo...
To Anup,
A rule of inference is a condition which tells when one can deduce a predicate from another one. So it can be seen as a kind of causality, but it occurs in the formal system of logic, and does not assume anything about the meaning of the predicates upon which it acts. On this point you can see any book about logic, or my book about mathematics in theoretical physics, which comprises a part on logic.
To Demetris,
I will not argue about the replacement of galilean geometry by relativist geometry, this is well beyond the scope of this thread. What I said is that science progresses by explaining scientific laws inside theories, these theories can be challenged, but they have always to account for what the replaced theory had explained. Relativist geometry does not invalidates galilean geometry, it shows in what circumstances galilean geometry is valid. And of course both can be checked. So a theory is a construct, it belongs to the realm of ideas, it does not pretend neither to be the truth (because it wants to be challenged), and nor to be reality (at best it provides an efficient representation of reality). The only way to build science, and to avoid the issues rised by philosohers such as Spinoza, Leibniz or Schopenhauer, is to accept the fact that there is a difference between reality (and truth) and science. A scientific theory is ephemeral, but it can be efficient, and this is what matters.
About cosmology : for a law, to be deemed scientific, one can be able to test the law in all possible configurations. And of course this is impossible for the universe in this totality, only God could do this, but then one enters into metaphysics (that is a narrative about statements that cannot be checked). This is the basic trick of Spinoza and Leibniz to explain that the evil can exist with a benevolend God : God has forcasted that the existent one is the best,,accounting for the constraints. But one cannot prove the existence of God. This is the classic kantian refutation.
Dear all,
I've recently come across the following book, which seems to be an important contemporary take on physics as mathematical formalism. The author is Max Tegmark, and he has been lecturing widely on the basis of the new book:
http://www.amazon.com/Our-Mathematical-Universe-Ultimate-Reality/dp/0307599809/ref=sr_1_1?s=books&ie=UTF8&qid=1412630390&sr=1-1&keywords=max+tegmark
Many of Tegmark's recent lectures are available on you tube. He is definitely a "think big" man, and a pronounced formalism seems to be at the heart of his approach.
H.G. Callaway
Dear all,
Here is a short quotation from Einstein concerning causality. It comes from Nature, 26 March 1927, p. 467. I quote from Eddington's note to his book, The Nature of the Physical World, in my new edition (just published in England),
http://www.cambridgescholars.com/arthur-s-eddington-the-nature-of-the-physical-world
See p. 292--Eddington's chapter devoted to Causation:
A few days after the course of lectures was completed, Einstein wrote in his message to the Newton Centenary: "It is only in quantum theory that Newton's differential method becomes inadequate, and indeed strict causality fails us. But the last word has not yet been said. May the spirit of Newton's method give us the power to restore unison between physical reality and the profoundest characteristic of Newton's teaching--strict causality."
Whatever we may think about the relationship between QM and causality--at most I take it that QM limits causality--it seems clear that in this passage Einstein sponsors causality. He doesn't think that his own work is inconsistent with causality. The idea that modern physics requires strict and rigorous formalism seems to have other sources.
H.G. Callaway
To all,
HG's remark is of importance. But, even if it is always a bit preposterous to interpret Einstein's thoughts, it seems that Einstein was thinking about determinism.
Causality and determinism are different concepts, and the latter is actually more important than the former.
In physics there are two kinds of phenomena.
1. Continuous phenomena. Then using their properties one can usually expect to extract the laws which govern the transition from one state to another, notably through differential equations, and this is the starting point to the converse : take the relations between the derivatives to forecast a later state. In this way one can discern some causality, however always with the restriction that there is no reason to priveledge one variable over the other..
2. Discontinuous phenomena. Then at best physics provide laws which are expressed through probability.
QM has accustomed physicists to see the probabilist formalism as a characteristic of reality itself. However, QM calulus can be seen as a way to provide efficiently solutions, without the need to assume any random behavior in the real world (on this point see my paper on QM revisited).
Even if non continuous phenomena are the rule rather than the exception, even in the macroscopic world, almost all can be explained as resulting of continuous phenomena at a smaller scale. However there are still some phenomena for which we have no satisfying answer, the transformation of elementary particles being the most important.
So, we do not have yet a clear answer about determinism, and obviously without this answer one can see that the issue of causality is a bit outpaced.
@Jean Claude
I would respectfully disagree with the last statement. A process can be both causal and indeterministic.
Consider a small example:
States={a,b,c,d,e}
Probabilities of transitions:
P(a,b) = 0.3
P(a,c) = 0.7
P(c,d) = 0.5
P(c,e) = 0.5
Occurence of (a) causes a transition to (b) or (c)
Occurence of (c) causes a transition to (d) or (e), all with the assumed probabilities...
The choices between {b,c} after (a) are indeterministic. Simillarly any of {d,e} may occur only after (c). The process is both causal (a is a cause for b,c and c is a cause for d, e) and indeterministic.
See also: http://www.informationphilosopher.com/freedom/indeterminism.html.
To Piotr,
1. Determinist means that there is some law which tells what happens in any given situation : if we know the initial state, we can compute the final state. And this assumes that there is no probability involved.
2. If you consider four states, you should consider all the possible occurences. So you can observe the occurences (a,b), (a,c),...Say that a is the "cause" of c is arbitrary : the only thing that you can tell is that there iis the probability 0.7 that both a and c are observed. For instance it is somewhat bizarre to say that a meson is the cause of the apparition of a pion, or some other particle. The cause that one looks after is the cause of the transition, and clearly for most if not all process represented by probability one cannot identify a specific cause, just combinations of initial and final states.
To Jean Claude (about your post on top of this page)
J.C. Dutailly":“A rule of inference is a condition which tells when one can deduce a predicate from another one. So it can be seen as a kind of causality, but it occurs in the formal system of logic, and does not assume anything about the meaning of the predicates upon which it acts. “ (End of quote)
The If…then …of logic is not temporal, because it is a logical truth that
If A then A.
Should we then say that it should be eliminated as an Axiom when logic is used in physics? I think not. In this case we would use only a fragment of logic. Otherwise we would have to say that something can be its own cause. And moreover it would have to be a logical truth that everything is its own cause. That would change the meaning of the word cause.
The logical solution is that “If A then A” says only things like “If it’s a Tuesday then it is a Tuesday”. It is not saying “A happens after A”. At most it says “If A is a cause then A is a cause”: - “If A then B” says then, that we will not have A without B, but not that A precedes B. It says things like “If it’s a Tuesday, then it is a weekday”. “If A then A” says only that everything is identical with itself.
Of course “If A then A” is not saying that everything is its own cause! It would only do when we decided to interpret the form in a temporal sense.
Cause/effect: They could be brought into a logical form with something like this: “There is a time point t1 and a time point t2 and t1 is before t2 and A is at t1 and B is at t2….” But this shows only a sequence, not causation. Logic does not tell us something about properties of objects and processes - but only about possibilities and impossibilities.
How can logic make sure that “If A then B” cannot be interpreted as cause and effect? That’s because A and B stand for SENTENCES, not for events or objects or properties. To be correct is also not saying “Tuesday > Tuesday” but “If the sentence “It is a Tuesday” (is true) then the sentence “It is a Tuesday” (is true).
This works with “Tuesday” and “weekday” because of the meanings of the words and also with “cause” and “effect” because of the meanings – but not with “force” and "acceleration” because it is not the meaning of the word “force” to “cause acceleration”. At this point the physical laws come into play. A physicist may say “When it causes no acceleration then it is no force”. (I don’t know whether a physicist would say that). Anyway that goes beyond logic. Would you agree?
To Martin
A rule of inference is different from a simple causality relation. It is used in the demonstration procedure (one side of logic).. Precisely it says that :
if A is a formula (supposed true)
and if (A=>B) is a formula (supposed true)
then B is a formula (supposed true)
It is slightly more complicated with predicates (using the operators for any..., there exists..)
Thus one can build demonstrations, by adding formula, starting with axioms, and the last line is the formula which is so proven.
This is pure logic, one never questions the meaning of A or B (which are formula, that is a predicate) : the only thing that matters is their value : true or false.
The commonly accepted definition of causality, given by Aristotle, clearly refers to a succession of events, with different cases (see Wikipedia). And in modern logic it is more muddled, with different definitions according to the authors. So, contrary to what philosophers used to do in the past, I do not see a great interest to this concept in Physics. What we need is a rule of inference, because this the crucial link in a mathematical demonstration.
I believe the objection to causality approach is possibly explained in jean claude Dutailly’s remark as, “if we need to look for causes to explain the effects, this quest is endless, and we must conclude at the impossibility of science, as an explanation of the real world.” For example, if we consider the fact that “when a force is applied on a mass it is accelerated” we may not be able to explain why it happens and therefore we may not be able to explain the cause for this happening. However, when we observe that a force is applied on a body we can definitely conclude that the effect of this cause viz., the application of force will produce an effect that is the body will be accelerated. May be the next day we will have some theory that will explain the cause for the above happening, but could be some more thing may remain unexplained. Limitation of our knowledge should not deny application of a concept that is otherwise capable of explaining physical happening and also follows our intuitive reasoning.
The cause and effect relation has been modeled using logic by C.A.R. Hoare (the Hoare logic), E. W. Dijkstra (Weakest precondition calculus), L. Lamport (Temporal logic of action). I have tried to model a state transition rule using cause and effect approach that uses temporal logic (Temporal Logic Related to Observation). One can see my paper entitled “Modeling of state transition rules and its application” which is available in the following link.
https://www.researchgate.net/publication/220631171_Modeling_of_state_transition_rules_and_its_application
Nondeterminism had been studied by E. W. Dijkstra in detail using logic in his famous book “A DISCIPLINE OF PROGRAMMING” and in many other papers. By inference rule on can derive theorems from the newly developed knowledge. This will be a continuous process. We will get new theories. Possibly a law will turnout to be a theorem but for this reason I do not understand why the application of a known methodology should be denied. We can accept that only when we get a logical contradiction.
Regards
Article Modeling of state transition rules and its application
Dear Jean Claude,
With inference rules we will not have axioms. But the formulas will have to correspond to complete sentences, when they are supposed to be true or false. Otherwise they will be open formulas (propositional functions) e.g. “something x is a circle” which is neither true nor false .I think so far we agree. Maybe we can also agree on a very general schema like
Language > logic > research > better language > logic > research > better language…etc.
“Better language” means that we use the words with more precise meanings.
Here comes the point where we do not agree:
You will have to decide: are formulas true/false - or are they predicates? First you write
“if A is a formula (supposed true)
and if (A=>B) is a formula (supposed true)
then B is a formula (supposed true)” (End of quotation)
Here the formulas have to be equivalent to complete sentences and cannot be predicates.
Later on you write:
“This is pure logic, one never questions the meaning of A or B (which are formula, that is a predicate) : the only thing that matters is their value : true or false.” (End of quote).
When a formula is supposed to be a predicate, then the form is e.g. “is a circle” or “x is a circle” This form cannot have a truth value “true” or “false”. In natural language e.g. “running fast” is neither true nor false as long as it is not attributed to a specific object and by this makes a complete sentence.
My point is that with logic we will not reach a point where we can attribute the predicate “is a cause” to an event. Maybe physics can. Why not logic? Because it acts on the level of complete sentences (closed formulas with bound variables) not on the level of single predicates (open formulas with unbound variables).
The complete sentence would have to be something like “x (= description plus space time coordinates) is an event” or “x (= desc plus…) is an event and y (= desc plus…) is an event and x is the cause for y…”.
The predicate “is a cause” does not follow from the predicate “is an event” but has to be written down before logic starts.
The sentences are like brackets. They encapsulate the predicates. They make sure that there are no properties without objects.
One can use the causality relation described by a suitable logic as described in my earlier post and the corresponding inference rule related to the logic to explain the behavior of some system. This is what is done to prove the correctness of a system defined by the state space approach.
Dear all,
The concept of causality is tied up with the arrow of time, since we hold that a cause must precede its effect. The arrow of time, in turn, is often thought to depend upon entropy and the second law of thermodynamics in particular. But there is some tendency in physics to think that the 2nd law is not "fundamental" and that all fundamental physical laws are reversible. For instance, Newton's laws allow one to predict the future positions of the planets in the solar system, but equally allow retro-diction of their prior positions, given the needed information about subsequent developments. The idea arises, in somewhat this way, that all of fundamental physics is indifferent to the arrow of time, and that we ought to think of events in terms of functions from one or another variable to another--in place of evoking causality with its temporal direction.
The physicist John Wheeler tells of asking Einstein about related points, late in Einstein's career, and reports the answer in his autobiography Wheeler 1998, Geons, Black Holes and Quantum Foam: A Life in Physics, pp. 166-167.
According to Wheeler, Einstein replied as follows:
I have always believed that electrodynamics is completely symmetric between events running forward and events running backward in time. There is nothing fundamental in the laws that makes things run only in one direction. The one-way flow of events that is observed is of statistical origin. It comes about because of the large number of particles in the universe that can interact with one another.
---end quotation
We have already seen that this view of Einstein's did not bring him to discount causality. He clearly evoked causality in his message to the 1927 Newton Centennial, quoted in a prior message. Subsequent emphasis on the reality of quantum randomness, would, I think, only serve to emphasize the reality of statistical considerations in physics. Effects which are significantly scattered, as in the familiar image of a glass of water falling off a table and shattering across the floor, are extremely unlikely to reverse themselves. Likewise, once we have a "collapse of the (deterministic) wave-function" the physical results are extremely unlikely to simply reform themselves.
The interest and viability of the concept of causality seems closely connected with the arrow of time and with determinations of predominate directions of development in physical systems. Those who emphasize the reversibility of fundamental physics, instead of being skeptical about causality, might do better to focus on how temporal direction and causality arise from the laws of physics.
H.G. Callaway
Dear all,
This thread is becoming more and more interesting...
1. It is not too difficult to show that the universe (at least locally) can be reprsented as a 4 dimensional manifold (we need 4 parameters to locate a point). But there is a fundamental symmetry brakdown (to use a common denomination) : one of the dimension is different. No observer can travel backward in time, and his present moves with him. It defines a folliation, specific to each observer. However two different observers never see objects moving in opposite direction of time. So one can safely assume that there is an "arrow of time", which has a universal meaning. It makes possible to speak of "before" and "after" an event in non ambiguous terms, and by this gives a meaning to causality (with always the same issue of identifying one special phenomenon as a cause).
2. In all practical physical experiments what we observe are really a succession of states (which occur in our successive presents). We state laws which specify the relation between the states (for instance PV=nRT means that the same relation occurs in the successive states, even if one of the variable has changed). From there, and for continuous process, one infers instantaneous laws, meaning relations which should be valid at a given time between the derivative of the variables defining the states. And it is a fact that all these instantaneous laws are symmetric (they should be valid if the direction of time was changed). However one can see that this symmetry is formal : there is no way to actually prove that this law, with an opposite direction of time, would be valid. Thus the conclusion should be that the true laws, scientific because they are falsifiable, should consider only successive states (just as the principle of least action), and if and when instantaneous laws can be deduced, one should say that only one of the two solutions should be kept. And it support the 2nd law of thermodynamics.
3. Moreover in all the examples of the 2nd law, there is discountinuous phenomenon involved (the broken glass), or a numbre of microsystems which interacts (the glass has its properties because of such interactions). So we should always consider the whole collection of microsystems, with their interactions (even if they are weak) during the experiment. Water (H2O) can be made with hydrogen and owygen, and dissociated, with exchange of energy and entropy, and the latter is involved because a mixture of hydrogen and oxygen gas has not the same ordered state as molecules of water.
The problem of non continuous procedures can be treated well if we use a discrete non spatial variable ('time') in order to descrive its evolution. It is not necessary to introduce a whole 4d manifold with very strict constraints in order to solve a problem.
The problem of causality is a big one, but here Anup has just asked a very specific question which returns us to the forgotten 'immediate action'...
To avoid possible contradictions one should formulate the energy-momentum variables (operators) in simultaneity with their conjugate partners (time and space). This would specifically avoid paradoxes without explicitly invoking causality.
in a given phenomenom, it is difficult to see the cause and its effect, because it may be a reversible interaction, but in experimental conditions, an inhibitor added to the observed system may inhibit the one direction reaction and not the other. In such experimental conditions the hypothetical cause and its effect may be in time-dependent order. Please, remember that I am an expert in biochemistry.
I was asked to help with an experiment that kept going wrong. I showed the investigator what was needed to achieve the desired result. Later, he told me that my method did not work. I asked how he performed each step to the method I proposed. When I observed that he had not done what I told him to do, he answered, I did what you meant for me to do.
That seems to be the gist of many of the answers to this question. The answers are to the question Anup meant to ask.
Should there exist a temporal ordering between a cause and its effect?
Marcel answered with, "I don't see how by definition the cause and the effect of the cause can occur at exactly the same time." Nor do I, Marcel.
There appears to be an argument that we cannot know the 'true' cause; therefore, how can we know the temporal ordering?
If a ball collides with another ball, the hit ball moves. The cause was the collision, the effect was movement. Aha! Something had to release the first ball, so that was the cause. No, no. The releasing only let gravity play its part, so gravity was the cause.
Well, maybe. The ball had to be lifted so it could be released. Lifting was the cause. But, but ... Somebody had to decide to lift the ball. The ball was lifted to be at one with the universe. Cause and effect are the same.
How silly! It is easy to see from the physical laws that cause must precede effect except when it is ambiguous. Look at equilibrium conditions. In order to change, everything at equilibrium must change at once. That makes it ambiguous. Which change is the effect and did each cause the other? Try doing that first.
Einstein showed that everything is relative, especially my cousin. If E = mc2, then from a space-time perspective and a time-space relative we must be one with the universe. Cause and effect are reversible at the same time, except in Texas.
There are social issues. Someone performs an act in order to cause something to happen. Since they wanted the effect, it must be the effect that caused the cause. That is perfectly understandable. We can prove it by logic.
A logical argument is necessary in a proof. It is integral to science. A logical proof requires that all parties accept the premises. Accepting nonsensical premises leads to a nonsense proof.
Here is the logical proof of the question:
Cause must precede effect.
A and only A causes B.
B occurs.
A came first.
This is the answer Anup meant me to find.
@Jean Claude
It seems that causal dependencies can be used in two directions:
- backward - to reason about possible causes of an observed phenomenon
- forward - to predict possible effects of a cause
I think, Computer Science is more concerned with the second. A common problem is to predict the future system behavior (e.g. a sequence of states) after making assertion on a current/initial state.
Moreover, very often an infinite behavior is analyzed (there is no final state). Typically, an expected behavior is expressed as safeness (something wrong will not happen ever) or liveness (something will eventually happen) properties.
In most cases, causal relations are quite well defined (at least at the level of compound processes). They reflect directly their specifications. However, the aggregated behavior is often nondeterministic due to involved concurrency and interactions with environment.
Model using probabilities are actually not used to reason about certain classes of systems, e.g. 0.01 probaility of catastrophic failure of a critical system worth 300M$ is not that promising (still 3M$ expected loss), however, they can be applied to assess performace, throughput, service level agreements, etc.
Dear Joseph L Alvarez ,
Normally the cause should precede the effect is obvious. However if we consider the very simple RC circuit as shown in the accompanying figure we find that if the switch is closed at t = 0 the input voltage vi(t) as applied to the circuit will look like a step function as shown. The output vo(t) is also shown in the figure. From the figure we find that the output starts building up at t=0, the instant when the input is applied. Starting of this state change is definitely an event and it happens due to the application of the input (cause). Apparently there is no “temporal ordering” between these two events. May be there is a catch – I am unable to find that.
'A logical proof requires that all parties accept the premises. Accepting nonsensical premises leads to a nonsense proof.' As I am just reading from Joseph Alvarez.
According the physics from the previous century, physics and psychology were complementary. The concept of cause and effect is already long forgotten. Also nonsensical is what the mind is and the mind without the unconscious has no value.
1. To Joseph
What I mean is that the concept of causality, which has mainy different definitions (see Wikipedia) is no longer necessary in Science. For centuries Philosophers, who have based their understanding of science upon causality, have shown that it leads to skepticism, and at least to the impossibility of efficient science. And the modern epistemology is not based upon this principle. If you want, it is generally possible to use it, because it is simple and practical from a pedagogical point of view, but it does not play a funding role.
2. To Piotr,
Your remark about information processing is good. Actually the idea of cause and effect appears in procedural softwares, which have been the bread and butter of programming for years. But now one looks for non procedural software, such as neuronal computing, where the idea of cause itself is fading. What we see are different equilibrium, resulting from the interactions of connected systems, And it is useless to look for a cause. In Physics one faces, almost always, interacting systems. The model of a single object, interacting with a single field (the equivalent of a procedural software) , is useful for pedagogical reason, but this is a simplification.
3. To Rita
"nonsensical is what the mind is and the mind without the unconscious has no value."
There is no knowledge, and certainly no scientific knowledge; without prior concepts which are formed in the brain, as abstract representation of phenomena. It has been fashionable, following Schopenhauer and Nietsche, and those who have (badly) copied them (Derrida, Deleuze, Foucault,...) to say that these concepts themselves come from somewhere (the social order, the unconscious,...) and so have no value. However in Science the concepts are defined with properties, and these properties themselves are linked with measure. So they have a link with the sensible world.(thiis characteristic make them distinct from metaphysical concepts). It is always possible to contest the concept, but when they are used in a scientific law, the law is valid if it can checked. And it is typical for the deconstructionists to look for a hidden cause (so to call for the cause and effect reasoning), to explain a law that they do not like. For instance in economics the concepts based on accounting are deemed "biased" to serve the interests of the capitalists, but whatever the critics, using these concepts it is possbible to understand how a company or a country becomes bankrupt. And this is what matters. And the deconstructionism is still popular only in some american universities.
Anup
A voltage is applied across the circuit. How long does it take to inform each element of the circuit? When does the first electron move? Effect follows closely upon cause. Field movement is part of the effect. Field movement is restricted to the speed of light. Cause precedes effect.
Since our celebrated theories like Newton's equations or quantum mechanics are time reversible, violations of time invariance has to be to be sought elsewhere.
What is the "cause" of this (e.g. the derivation of the second law in thermodynamics from statistical mechanics) is still an open problem. Since gravitation is not yet unified with quantum mechanics, the latter does not offer rigorous proof of the one- directedness of time etc. although this can be most likely asserted.
So, dear jean, what is the current 'fashion views' in recent epistemology, we have put away causality and we replaced it with what? interdependence of Lernean Hydra kind?
http://en.wikipedia.org/wiki/Lernaean_Hydra
The laws of physics (those simplified descriptions of the world) are often discussed without regard to cause and effect. "The center of mass of X is accelerating ..." No need to discuss the force. No need to mention how much mass. It is particularly useful to avoid discussion of that mass in the interval between at-rest and accelerating. It is nice to avoid that messy situation of force turning into momentum.
When my daughter went through the stage of asking why, she was not comforted by having a physicist father. I hugged her a lot to compensate for that.
The ability to discuss the world without introducing cause and effect does not preclude nor compensate for discussion of cause and effect. It also does not change the temporal relationship.
Dear all,
Briefly, regarding the famed Humean skepticism on causality, though it has had many supporters, I think this is pretty clearly overdone. As a general matter, the skeptical perspective built into Hume's philosophical and epistemological premises are so strong, that he can derive skepticism about just about any topic one might care to consider. In consequence, his skepticism concerning causality is no great surprise. What significance does it have?
Where attention may be needed is on Hume's concept of "custom." How are we to understand the "custom" which allows Hume to recover a semblance of ordinary or philosophical talk about causality? Is this purely social, or even social-political? Is it purely conventional, or do factual and theoretical developments plausibly enter into it? If so, then it no longer seems purely conventional, and the idea of a purely social (or political!) structuring of the custom upon which judgments of causality depend becomes cognitively implausible. The so-called "grue" paradox --including the many responses to it--seems to be of interest here. More generally, I think we need to recognize that judgments of relevancy (relevant causal conditions) are of interest regarding judgments about causality, and I just want to briefly suggest that these are, in fact, deeply embedded in the specific developments of particular sciences and scholarly disciplines. In this way, highly formalized approaches become excessively conservative, in relation to the advance, or potential advances, of particular sciences or scholarly disciplines--at the risk of closing themselves off into dogmas based on pre-established findings or discoveries. In contrast with this, we need to take full advantage of the perspective provided by the history of a branch of inquiry--in order to bring it forward.
How indeed are we to understand the history and eclipse of "continental rationalism" in Western thought? May I suggest that this was a closing off among those sharing the same "a priori" intuitions? But the cure for excessive dominance of particular intuitions, when this blocks inquiry, and excludes the critics (one may expect ulterior motives), is to turn back in the direction of, empiricism, experiment and observation. The mystique of Human skepticism is that it appears to combine empiricism with a central emphasis upon "custom" in the validation of judgments of causality.
I don't know if this is the best place to examine related questions, but I suspect the themes are running in the background.
H.G. Callaway
Dear H.G.
It is clear that we are facing the old schism between anglo-saxon empirism and continental rationalism, and of course I find that you are a bit too quick to announce "the eclipse of continental rationalism". I stick to my guns, and will defend strongly Kant and Popper against Hume, Berkeley and the others ! We have now a sensible epistemology to account for the modern science, and to Demetris, this is nothing like the Lemenean Hydra. Without being too bold, but to save time, those who are interested can find some explanations in my paper on Common strutures. These are not my personal findings, just a summary of what I have understood as to be the present day consistent epistemology, regarding science. I can see that the cause and effect principles can be of importance in the empirism thought, because it is so closely linked to primarry (I would say primitive) perceptions (even with a bit of measure added). But as usual with empirism it leads only to an endless sequence of questions. For a good reason : a new theory can never be borned from the simple observation of facts, how sophisticated it is. Theories are borned in the brain of scientists, and built from concepts.
To Rita,
I had try to answer your remark. Perhaps it was a bit out of the scope of your thought. There is no doubt that concepts are borned in the brain, and the complicated mechanisms at works there can be seen as complementary to the birth of a theory. However, in a cartesian way, I would leave this issue aside, as long as we do not know more about them. As I have said before, Science does not pretend to achieve truth and full knowledge of reality, just to provide an efficient way to act on the world. For me it suffices. Even a bent tool can still be useful.
Dear Dutailly,
Many thanks for your thoughtful and suggestive reply. My comment on "the eclipse of continental rationalism" stands. I mentioned this in passing, though, thinking primarily of the eclipse of classical modern rationalism. I am aware, of course, that Popper uses the term "rationalism," but I think of Kant as belonging to the old early modern rationalism.
Your image of empiricism is of interest here, connecting this, as you do with "primary" or "primitive" perceptions. I think that this image of empiricism is antiquated and would surprise many or most empiricists of today --and those conducting science in an empirical spirit.
It belongs to empiricism, as it is generally understood, I submit, that even the smallest, or most accidental observation can, in principle, bring down the most established theory, and even perhaps invalidate or put in question the most cherished concepts and methods. Empiricism, as it is best understood is no a matter of the origin or source of concepts, it is instead a matter of their needing to stand the test of empirical scrutiny, in terms of the predictions (or lack thereof) arising from theories employing particular concepts.
It seems to me clear, in any case, that there is no end to inquiry on an empiricist conception, no final and ever valid theory of everything. But this strikes me as a side issue--in the present context. I can imagine that others may see the matter differently. Personally, and philosophically, though, I see no problem in the prospect of an endless sequence of questions.
Consider then the Newtonian conception of absolute time: a single ticking clock of the entire universe, which would allow for simultaneity of events no matter their location or relations. That is a very appealing concept for many, and was long accepted along with Newtonian physics. But as Newtonian physics fell, in light of contrary evidence, so did this concept fall. Though theories are not always or often born from a simple observation of facts, observation of facts can knock them down, along with embedded concepts. That is all that is needed for a viable empiricism. I don't see that talk of concepts is going to get us a new-fangled rationalism. I suspect that even the most formally inclined would be surprised at this idea.
H.G. Callaway
Dear H.G. (I do not know of your first name...)
Of course I fully agree with the role of experiment, as emphasized in empiricism (sorry for the mispelling), as an essential and critical part of any scientific endeavour. This is what distinguishes science from metaphysics. But, and perhaps on this point I extend somewhat the usual vision of modern epistemology, and this is a product of this kind of discussion, I feel that the distinction between scientific laws (with the necessary concepts, properties and relation with measure) on one hand, and theory, that is the construct of a consistent and formalized set of fundamental assumptions and laws, on the other hand, as useful. Laws are closely related to experiments, and even if they require, to be clearly stated, concepts, they do not suffice. This was the main limitation of Descartes' picture. Science requires more, in the quest to understand "what it is" (das dasein), that is to find fewer but more fundamental,laws. Actually the laws are not refuted, they are only relativised (as Bachelard explained). The issue of epistemology is then to explain the rules in the control of the chain of successive theories. The first example which comes to mind is the place of QM : the fact that its place is still discussed, and uncertain, leads me to find a rational explanation, and I think that I have achieved this, showing that QM (in its basic form of the axioms) is actually not a physical theory at all, but more a mathematical consequence of the formlism used in models. But on the same topic one can look at the hierarchy of science (à la A.Comte). Even if biology comes from chemistry, and thus from physics, we still have a gap in explaining the meaning of life. And even Mathematics has not the status that it seems. We have, and this is a fantastic property of the humain brain, the capability to find, intuitively, the right axioms to find useful theories : the theory of sets could work for most usages, without the axiom of infinity, that we have added by convenience, because if is useful. Notice that this is an axiom which cannot be related to any physical evidence. By doing so we know, from the Godel's theorem, that infinitely many other axioms could be added: are they useful, could they become useful some day ? Who knows ? In this way Mathematics becomes a science, open to discovery.
Dear Dutailly,
You'll forgive me, I trust, in finding your reply somewhat underwhelming --and unpersuasive. The idea that there are substantial "rules in the control of the chain of successive theories," is simply historicism --a thousand times refuted and the last refuge of the history of Western rationalisms. At the very least, you've failed to distinguish your position from historicism. See, e.g., Popper (1957) The Poverty of Historicism.
H.G. Callaway
Dear H.G.
There is no disagreement about the role of experiment in checking a scientific law : it must be always be verified whatever the circumstances, and this is this universality which provides their power. Falsifiability is at the core of Popper's epistemology. However Popper was not a true scientist and dedicated a large part of his work to political science, notably to the refutation of marxism. I am not a philosopher, but I guess that his vision of historicism was somewhat influenced by Hegel. My question about the refutation of theories is not based on the same premices, I do not see any dialectic at work in this : a theory is a way to assemble together several laws related to the same field. We have a theory of gravitation, of electromagnetism, the atomic theory in Chemistry, genetics in biology,...Actually what scientists do is not just to refute laws, but to refute theories, and they do this by using the background provided by theories to invent new laws that can be tested. The precession of the orbit of Mercury is a fact that contradicts Kepler's laws, and the usual Newton's theory. Einstein used general relativity to see how it would be possible to explain the phenomenon, so he postulated a new law, and its experimental proof has been crucial to support General Relativity. It is clear that this new law, which is in agreement with the measures, could not come from observation only. General Relativity was not limited to the revision of a law about Mercury, it obliges to revise our representation of the world itself, and this is what makes a wonderful achievement. All this has been explained by Bachelard. This process, at work in the progress of science, is a fact, but it needs to be better understood
You say :
"In contrast with this, we need to take full advantage of the perspective provided by the history of a branch of inquiry--in order to bring it forward."
so,to refer to the history of science (how ?) would be better than to understand how the refutation of theories works, and try to formulate some rules in order to put some order ?
A clear consequence that I see is QM : as Feynman and others have stated :"nobody can claim to understand it", but, as far as the computation works, who cares ? This certainly can satisfy empiricists, not me. Many physicists are pleased with the standard model, which dates back to 1973, and were glad to see the proof of the existence of the Higgs boson. But, as any scientist can tell, the standard model is an awful bric a brac (that you can admire on Wikipedia) and I think that no more than 1000 scientists in the world, almost all professionally engaged with the project, can genuinely understand what the proof means. If we do not try to check the checkers, the progress of science will be left to the "publish or perish" and the endless struggle for fame. I prefer the rules of the brain than the rules of the media.
Kant said "wage kennen !" dare the knowledge, and Hilbert's motto has still some value : "wir mussen kennen, wir werden kennen". And knowledge is not the accumulation of well checked laws.
Dear Dutailly,
Historicism needn't involve dialectics, though it often does. The refutation is nothing purely political. See the actual arguments.
http://www.amazon.fr/gp/offer-listing/0710046162/ref=tmm_other_meta_binding_used_olp_sr?ie=UTF8&condition=used&sr=8-3&qid=1412827114
In any case, this last posting is much better, I believe. Lets keep this from becoming merely a dialogue of two.
H.G. Callaway
Photons that have been emitted from the cosmological microwave background billion of years ago can enter a detection apparatus whose interaction instantly determine the state of this photon at the time of its emission billion years ago.
Quantum Theory and Measurement
Edited by
John Archibald Wheeler
and
Wojciech Hubert Zurek
http://www.forizslaszlo.com/tudomany/wheeler_law_without_law.pdf
Spacetime, can this not mean that time has a space dimension. So not the inversion of time, but inversion of spacetime. although the parity symmetry proposed by Pauli was not proved by the experiment.of Lee and Yang as I could read.
Dear Anup,
answering your question would require a clear definition of the pair cause/effect, which, depending on tastes, may be more or less easy task. Just to be clear, let me try to explain with an example. Would we like to consider the minimum action principle as a possible cause ? Of course there is nothing wrong in considering it or not. It is just matter of definition, and definitions may be useful or useless. Never true or false. Even if we constrain the category cause/effect to events, things do not become simpler.
A part these general difficulties, your example is probably not the best one could use to illustrate the point. One of the reasons is that a clear understanding of the principles of mechanics is less widespread one could think. Even in this discussion, I have seen opinions which completely ignore decades of discussions and critical analysis of the Newton's laws of motion.
In that respect, I would just say that asking if the force is the cause of the acceleration or if the presence of an acceleration is the cause of speaking about a force can considered equivalent by anybody who read Mach's critique of Newton or post-machian discussions.
Instead, what is related with the cause-effect idea is the question if the evolution at t>t0 depends (and how) on the configuration of the system at time t0. Or if the acceleration of body 1 in a system of particles depends on the position of body 2 at the same time or at a previous time. Or even by the configuration of a field at the same time...
You see that the answer can vary a lot depending on the definition one would like to privilege for "cause". What is yours ? You mentioned "dependence". But one could have dependence even with respect to future events (e.g. final dependence). So, it is im[possible to say something without agreeing on the definitions we want to use.
Causes and effects are typically related to changes, events, or processes. Anything that affects an effect is a factor of that effect. In science, an experiment conducted with controlled factors including objects, processes, properties, variables, facts, and states of affairs, there may exist an obvious temporal ordering between a cause and its effect. However, natural phenomena taking place with so many aspects beyond control and unrecognized causes are complicated and intertwined, often overlapping thus making it difficult to establish a natural order between cause and effect.
Dear jean, you will never find what the initial H.G. mean! I have tried many times by myself! :)
As for the schools of thought: I am open to accept a kind of thinking with many degrees of freedom, but, despite my willing, I am a being of a specific local universe and thus I am constrained to follow a narrow band route. So, although in principle every possible math formulation could exist, in our local universe probably does not apply. Thus is our curse: we can imagine how things could be, but we have to find our grounded reality.
There is one standard of truth in science, the world. The cause of an effect might be completely shrouded by multiple layers of interaction, each having an interaction with the cause and each producing an effect. The question is if A caused B. The action A on B may entail one or more chaotic systems. C may also cause B. The inability to discover the relationship between A and B does not alter the temporal relationship of A and B.
Science attempts to simplify cause and effect to the point that effect may be idealized in a manner that secondary effects and contributing causes are ignored. The ideal might be elevated to the status of a law. The law is completely correct in the ideal but is not correct in the world. The laws are what we teach. The world is where we learn by applying the laws. There is the suggestion that pure science is only in the brain. We cannot think of science without the laws of science. We cannot refute the laws without reference to the laws. We cannot develop beyond the laws or a refutation of a law without first mastering the laws. We cannot master the laws without occasionally touching the ground.
- Badges
- Science method
- Similar topics
- Philosophy