Stephen Hawking's positivism

  • Thread starter John Jones
  • Start date
In The Universe in a Nutshell (p. 31) Stephen Hawking describes his science positivism:

.… If one takes the positivist position, as I do, one cannot say what time actually is. All one can do is describe what has been found to be a very good mathematical model for time


What has Hawking said here? Hawking invokes a grand metaphysical project of what things really are - "what time actually is" but with other science/logical positivists we must presume, wants to deny metaphysics and ontology. This he does on the basis that he is "unable to say" what the particulars of a metaphysical project (like Time) are.

And what is the role of mathematics in this Hawkinian idiosyncratic venture? If, with other positivists, he wishes to deny metaphysics and ontology in mathematics and the world, then how does he know what a given calculus is meant to be about?

Hawking's attempt at positivism fails. He attempts to justify, piecemeal, the positivist denial of metaphysics/ontology by privileging its denial over its affirmation in the proposal that "one cannot say what [X] really is" (where X is a metaphysics, in this case a metaphysical description of Time). Derrideans might pick up on this logic of the "priviliged binary".

Hawking's appeal to mathematics as a player in his positivism misses the mark. It may even work against him. For mathematical syntax gives an incomplete description of mathematics, yet it is the syntax alone that is set up as a sufficient condition of it. The stops and starts of syntactical manipulation are not themselves syntactically engineered. We need a non-syntactical manoevure to identify and manifest, or commit syntactical changes, and metaphysics is one such manoevure. For example, I, non-mathematically, decide when a calculation is complete, and what to calculate. Here I can allude to the Tractatarian project as a whole, where Wittgenstein continually draws attention to the limits of language and logic - where syntax is identified and manifested by the "ineffable" template (later re-emerging as language games in Philosophical Investigations). Kant had a similar project going in transcendental idealism, where no objects are to be had without their identifying and manifesting conditions. Yet, for Hawking's transcendental realism, as for other positivists, we see the denial of the manifesting conditions of a syntactical world. Hawking's positivist venture is twofold: incoherent, and a picture of a world of unidentifiable, syntactical rubble.
 
Last edited:

apeiron

Gold Member
1,971
1
Hawking's positivist venture is twofold: incoherent, and a picture of a world of unidentifiable, syntactical rubble.
I'm no particular fan of Hawking's musings on epistemology, however he still takes the untroubling position that modelling involves both formal descriptions and a process of measurement. So this criticism seems unjustified.

How the "syntax" of the description arises is irrelevant to the success of a model. Success here hinges on how well the theory predicts what will be measured in practice. And so models either work or they don't.

How the syntactical operations of mathematics arise is an interesting question of course. But another question so far as modelling is concerned.

The right question to ask of science is whether it is indeed making all the measurements it could, or whether it is ruling certain things out because they "don't fit the theory". And metaphysical speculation is the obvious way to highlight the kinds of things that are not being modelled or measured by current "well working theory".

So positivism of the "shut up and calculate" school is narrowminded.

But modelling relations theory - here I am thinking of Rosen, Hertz, Pattee - does stress that modelling involves both formal theory and informal measurements. A degree of arbitrariness is accepted in the epistemology as inevitable. Yes, decisions get made about truncating measurements. There are parts of modelling that are (formally!) unformalisable. But which are then justified pragmatically. They are habits or practices which over time prove themselves to work.

So you appear to be arguing some purist's case. And modelling epistemology does not in fact demand this formal purity. Thus Hawking's positivist rhetoric is not inherently self-contradicting.

It might be read as another unnecessary example of philosophy-bashing. I don't know, but this might be what you are reacting to? But even then, I would have to agree that the success of existing physical models have to act as a constraint on modern philosophy. They narrow the metaphysical options and questions in pretty radical fashion.
 
182
1
Hawking's appeal to mathematics as a player in his positivism misses the mark. It may even work against him. For mathematical syntax gives an incomplete description of mathematics, yet it is the syntax alone that is set up as a sufficient condition of it. The stops and starts of syntactical manipulation are not themselves syntactically engineered. We need a non-syntactical manoevure to identify and manifest, or commit syntactical changes, and metaphysics is one such manoevure.
We do? Why?


For example, I, non-mathematically, decide when a calculation is complete, and what to calculate. Here I can allude to the Tractatarian project as a whole, where Wittgenstein continually draws attention to the limits of language and logic - where syntax is identified and manifested by the "ineffable" template (later re-emerging as language games in Philosophical Investigations).
I don't understand the problem. Rules cannot account for themselves. Are you saying that they should in order for Hawking to make the claim?
 
I'm no particular fan of Hawking's musings on epistemology, however he still takes the untroubling position that modelling involves both formal descriptions and a process of measurement. So this criticism seems unjustified.

How the "syntax" of the description arises is irrelevant to the success of a model. Success here hinges on how well the theory predicts what will be measured in practice. And so models either work or they don't.

How the syntactical operations of mathematics arise is an interesting question of course. But another question so far as modelling is concerned.

The right question to ask of science is whether it is indeed making all the measurements it could, or whether it is ruling certain things out because they "don't fit the theory". And metaphysical speculation is the obvious way to highlight the kinds of things that are not being modelled or measured by current "well working theory".

So positivism of the "shut up and calculate" school is narrowminded.

But modelling relations theory - here I am thinking of Rosen, Hertz, Pattee - does stress that modelling involves both formal theory and informal measurements. A degree of arbitrariness is accepted in the epistemology as inevitable. Yes, decisions get made about truncating measurements. There are parts of modelling that are (formally!) unformalisable. But which are then justified pragmatically. They are habits or practices which over time prove themselves to work.

So you appear to be arguing some purist's case. And modelling epistemology does not in fact demand this formal purity. Thus Hawking's positivist rhetoric is not inherently self-contradicting.

It might be read as another unnecessary example of philosophy-bashing. I don't know, but this might be what you are reacting to? But even then, I would have to agree that the success of existing physical models have to act as a constraint on modern philosophy. They narrow the metaphysical options and questions in pretty radical fashion.
The positivist position eschews metaphysics and ontology. Yet Hawking appeals to an idea of what a thing, like Time, "actually is". This looks like an ontological or metaphysical speculation, and intimates a contradiction in his positivist sympathies.

The fact that Hawking i) cannot say what Time "actually is", and ii) says that Time is represented in mathematics, doesn't amount to a denial of the role of metaphysics and ontology, but again, seems to throw a line in its direction, again intimating a contradiction in his positivist position.
...

That brought me to my main second point which was that mathematics and science - whatever the success or method of application of their models or theories - cannot operate - have no ordered syntax- without ontology and metaphysics as these are two conditions that define the limits, hence identify, mathematical and scientific syntax. By scientific syntax I mean the objects of science (gravity, the electron, etc). For example, without some form of identifying, or transcendental condition (like Witts language games or Kant's categories) science would not be able to distinguish the TV from the carpet it stands on, or the electron from empty space. This is why I used the term "syntactical rubble".

This problem - the problem of transcendental realism's inability to postulate objects, runs through any positivist venture. But there are other reasons why Hawking's position looks awry, as I first mentioned ( i) and ii), above).
 
Last edited:
We do? Why?


I don't understand the problem. Rules cannot account for themselves. Are you saying that they should in order for Hawking to make the claim?
The multiquote button is not working.


Logic and mathematics require identifying conditions for their syntax. Ontological conditions define the behaviours of their signs/syntax. For example, signs don't vanish and appear without redress. This behaviour is based on the ontology of the physical object.

We also need identifying or transcendental conditions for all objects of science. There are no boundaries in nature. The boundaries that identify Suns and electrons are delineations that we make on the manifold of nature. Unlike cartoons, which sometimes have outlines drawn on their figures, there are no such delineations in nature itself.

If Hawking wants his positivism to work then he needs to first kickstart it with a delineator, or tanscendental condition. If we don't supply it ourselves through ontology and metaphysics then the objects, rules and syntax themselves must supply it. Although I hadn't said it, this is animism.
 
Last edited:
182
1
Logic and mathematics require identifying conditions for their syntax. Ontological conditions define the behaviours of their signs/syntax. For example, signs don't vanish and appear without redress. This behaviour is based on the ontology of the physical object.
The identifying conditions are in the axioms.

We also need identifying or transcendental conditions for all objects of science.
Identifying with reality yes, but what transcendental conditions?

If Hawking wants his positivism to work then he needs to first kickstart it with a delineator, or tanscendental condition.
You mean a God, as in something more than what can be said meaningfully?
Although I hadn't said it, this is animism.
That's unfortunate.
 
JJLogic and mathematics require identifying conditions for their syntax. Ontological conditions define the behaviours of their signs/syntax. For example, signs don't vanish and appear without redress. This behaviour is based on the ontology of the physical object.
The identifying conditions are in the axioms.
The axioms tell us how syntax is to behave. This is an ontological description. But we also need metaphysics to intrude on math syntax. Otherwise, we cannot tell where a calculus is to start or end, or what it is about.

JJ We also need identifying or transcendental conditions for all objects of science.
Identifying with reality yes, but what transcendental conditions?
See my TV example.

JJIf Hawking wants his positivism to work then he needs to first kickstart it with a delineator, or tanscendental condition.
You mean a God, as in something more than what can be said meaningfully? That's unfortunate.
No, nothing to do with God. Hawking needs objects, like all of us. So, if we do not set the conditions that describe the physical limits of these objects (careful now, see my TV example again), then the objects must set their own limits. But how?

(This is very awkward- The multiquote button is not working.)
 

apeiron

Gold Member
1,971
1
The positivist position eschews metaphysics and ontology. Yet Hawking appeals to an idea of what a thing, like Time, "actually is". This looks like an ontological or metaphysical speculation, and intimates a contradiction in his positivist sympathies.
I'm still not getting any clear idea of what there is to complain about here.

Why attack Hawking's populist statements? They can be taken as a bit of banner waving and even he would have a more nuanced position if questioned.

So are you attacking positivism in general (and then which of its many variants?), or just Hawking's public statements (which may indeed be self-contradictory)?

It seems plain enough to me that all Hawking is committing himself to here is that modelling reality involves formal theories (syntactical in that they are expressed as closed stories of causal entailment) and then the informal business of measuring, observing, predicting (which would include also the abductive leaps that are the "metaphysics" and axiom forming that are needed to get any formal theory off the ground in the first place).

So the model presumes there is a reality to be measured, and specifies the general nature of these measurements (as the kind of data which the model can manipulate). The fact that it works then shows that the theory is "true" - or fit for the subjective/epistemic purpose for which it was designed.

Positivism can sometimes pretend to eschew metaphysics, but that is again just banner-waving, a symptom of a moment in social history. We all know that abductive leaps are needed to get any modelling game going.

The fact that Hawking i) cannot say what Time "actually is", and ii) says that Time is represented in mathematics, doesn't amount to a denial of the role of metaphysics and ontology, but again, seems to throw a line in its direction, again intimating a contradiction in his positivist position.
Time becomes what we measure it to be. In relativity, it becomes a space-like dimension (and so gains the improbable feature of reversibility). In thermodynamics, it becomes an entropic gradient (and so gains a more realistic irreversibility).

A really strict positivist goes too far because he will say the model is all we can know - and no further speculation is allowed about the true nature of things. Yet in practice, you still have to make fresh abductive leaps of thought to generate fresh hypotheses about nature.

And in practice, most positivists who actually achieve anything new, will have had to make such leaps. It is only those who refuse ever to think outside the box who are doomed to remain forever within their self-made box.

This problem - the problem of transcendental realism's inability to postulate objects, runs through any positivist venture.
This is what I don't get. Where is the evidence that positivists don't postulate objects (or entities, or processes)?

Are you saying that they don't legitimate their actions at the level of their epistemological assertions? Which may be true, I haven't really worried about that issue much. Or that they don't postulate objects in practice when they come up with their useful theories?
 
I'm still not getting any clear idea of what there is to complain about here.

Why attack Hawking's populist statements? They can be taken as a bit of banner waving and even he would have a more nuanced position if questioned.

So are you attacking positivism in general (and then which of its many variants?), or just Hawking's public statements (which may indeed be self-contradictory)?

It seems plain enough to me that all Hawking is committing himself to here is that modelling reality involves formal theories (syntactical in that they are expressed as closed stories of causal entailment) and then the informal business of measuring, observing, predicting (which would include also the abductive leaps that are the "metaphysics" and axiom forming that are needed to get any formal theory off the ground in the first place).

So the model presumes there is a reality to be measured, and specifies the general nature of these measurements (as the kind of data which the model can manipulate). The fact that it works then shows that the theory is "true" - or fit for the subjective/epistemic purpose for which it was designed.

Positivism can sometimes pretend to eschew metaphysics, but that is again just banner-waving, a symptom of a moment in social history. We all know that abductive leaps are needed to get any modelling game going.



Time becomes what we measure it to be. In relativity, it becomes a space-like dimension (and so gains the improbable feature of reversibility). In thermodynamics, it becomes an entropic gradient (and so gains a more realistic irreversibility).

A really strict positivist goes too far because he will say the model is all we can know - and no further speculation is allowed about the true nature of things. Yet in practice, you still have to make fresh abductive leaps of thought to generate fresh hypotheses about nature.

And in practice, most positivists who actually achieve anything new, will have had to make such leaps. It is only those who refuse ever to think outside the box who are doomed to remain forever within their self-made box.



This is what I don't get. Where is the evidence that positivists don't postulate objects (or entities, or processes)?

Are you saying that they don't legitimate their actions at the level of their epistemological assertions? Which may be true, I haven't really worried about that issue much. Or that they don't postulate objects in practice when they come up with their useful theories?
I needed a quote, as per the rules of the forum, so I used Hawking's. I then tackled Hawking's positivism, before making a general point about positivism.

Models of reality generally appear as evocative literature that helps to make a metaphysically neutral mathematics more palatable and give it practical direction. Though, against science, I think the gulf between them is unbridgeable.

Time is a metaphysical concept in all its popular and esoteric forms, for it isn't evidential.

Positivists can't postulate objects because positivists won't appeal to ontology and metaphysics. Ontology describes object behaviours, and metaphysics sets up the limits and types of these objects. Without these behaviours and limits all we have is colourless syntax or rubble, as I called it.

Metaphysics puts life and direction into mathematical syntax, while ontology tells its syntax how to behave.
 

apeiron

Gold Member
1,971
1
Positivists can't postulate objects because positivists won't appeal to ontology and metaphysics. Ontology describes object behaviours, and metaphysics sets up the limits and types of these objects. Without these behaviours and limits all we have is colourless syntax or rubble, as I called it.

Metaphysics puts life and direction into mathematical syntax, while ontology tells its syntax how to behave.
OK, Hawking is just a good example of an attitude. Agreed on that score.

But positivism is also possibly a too broad target. As it comes in many stripes. Yet we perhaps agree that when positivism pretends it does not need some grounding in metaphysics or ontological speculation, then it is either fooling itself, or trying to fool us.

This is not merely dishonest, but unsound because epistemology needs to deal with the process of grounding models in some systematic fashion. It is part of the exercise that has to be spoken about and justified. We must know the reasons for choosing to do things a particular way.

I think the quantity~quality dichotomy is useful here. All scientific modelling has to postulate qualities (semantic abstractions rather than ontic objects) that then become the measured (the quantified).

So this puts things satisfactorily back in the realm of the irreducibly subjective. It is not about what is "out there" but a notion about a quality of my experience which seems suitably general as an idea.

An example of such qualities in physics are energy, matter, space, time. They seem natural enough categories of experience - even if ineffable on closer examination. And then science can start doing measurements in the name of these qualities. Eventually, there may seem only to be the objective act of measurement left because the original quality, such as time or energy, has become so generalised as to have lost all particular semantic reference. There is only the web of syntactical operations and they seem to refer to nothing because the words - time or energy - no longer seem to mean anything very clear.

At this point, if you break up that syntactic web to attempt to recover the buried metaphysical semantics, then you will seem to be left with nothing except your rubble of syntax.

And if you try to think about time or energy - the meaning behind the words again - a positivist will accuse you of stepping outside the syntactical framework that seems all that is necessary to him.

The scientific project is to generalise the subjective qualities of experience to the point that they are barely there, and everything else becomes reduced to the act of quantification, making the bare measurements needed to animate the quality-based equations.

You end up with F=ma. An instruction kit on what has to be measured, but no qualitative understanding of why this relationship exists.

So is this the source of the tension? All science has to be grounded in subjectivity, in metaphysics. But science is seen as a project to reduce qualitative impressions (semantic concepts) to quantitative measurements (syntactical operations). So it wants to be seen as rejecting its own origins. Hence the rhetoric we hear from positivists.

[EDIT] The same could be said in spades for mathematicians of course as they would prefer to believe they enjoy a hotline to a rational Platonia rather than admit that all their ideas are derived from subjective interactions with the world.
 
182
1
The axioms tell us how syntax is to behave. This is an ontological description. But we also need metaphysics to intrude on math syntax.
Again, why do we need metaphysics to be introduced into math syntax? I'm not sure of what kind of metaphysics are you asking for...
No, nothing to do with God. Hawking needs objects, like all of us.
Time isn't an object. And now you're saying that Hawking needs an object (like what?) to describe time? The mathematical models do that for him.
Hawking said:
If one takes the positivist position, as I do, one cannot say what time actually is.
 
Last edited:

Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving

Hot Threads

Top