by Neil Rickert

I skipped posting last week. I had planned to post about knowledge and belief, but decided to skip that post.

Conventionalism is interesting, in part because much of our life seems to depend on social conventions. And, in part, because philosophers seem to be strongly opposed.

According to Wikipedia, “Conventionalism is the philosophical attitude that fundamental principles of a certain kind are grounded on (explicit or implicit) agreements in society, rather than on external reality.”


It is usually agreed that a social convention is an agreement, perhaps implicit rather than explicit.

The rule that we should drive on the right side of road is often mentioned as an example of a convention. In some parts of the world, including Australia (where I grew up), people instead drive on the left side of the road. That there was a choice between driving on the left, or driving on the right, illustrates why conventions are said to depend on arbitrary choices. But those two choices (left vs. right) are not the only options. For example, there could be a system where people drive on the left on even numbered days and on the right on odd numbered days. This would be more confusing, with probably more accidents. But it serves to illustrate that there is often a degree of pragmatism in our choice of convention. Saying that a convention is an arbitrary choice does not rule out the involvement of pragmatism in the making of that choice.

Poincare proposed conventionalism for geometry. In his view, the axioms of geometry derive from our measuring conventions. I agree with Poincare on that.

Hilary Putnam argued against conventionalism in “The Refutation of Conventionalism”. One of his arguments was that under conventionalism there could be no matters of fact. I just measured the height of my desk as 74 cm. That’s a matter of fact which depends on the measuring conventions which define the centimeter. From the way that I look at it, all facts are relative to the conventions that we follow when observing those facts.

Several years ago I suggested to John Wilkins (on his blog), that biological species are conventional. His response was that this would make the species concept an empty one. I disagreed, but the discussion did not go very far.

Newton’s laws

I won’t discuss species in this post. Perhaps I’ll return to that in the future. Today, I want to argue that Newton’s laws were conventions.

Looking at that definition from Wikipedia, quoted above, Newton’s laws met the requirement of being fundamental principles of science, and of depending on agreements (that scientists agreed to follow them). But were they grounded on those agreements rather than on reality? In my view, they were.

People tend to look at Newton’s laws and explain them as based on observation and inductive generalization. But that seems mistaken.

Imagine that time traveling were possible, and that we could take Newton’s laws in a time ship back to when we could have discussed them with Aristotle. It seems to me that Aristotle would have seen Newton’s first law as obviously wrong. To Aristotle, motion involved moving things slowing down by their nature. And Newton’s first law contradicts this.

Newton’s second law, f=ma, would have been a mystery to Aristotle. He would have heard of weight, but not of mass. He probably would have thought of force in terms of when we push against something, but he would not have contemplated a force of friction.

No, Newton’s laws are not grounded in what would have been observed. Newton’s laws were a package deal. They gave us a whole new way of describing aspects of reality. And Newton’s laws are the conventions that define this new way of describing. The concepts of motion and force are changed from what Aristotle would have used. And even acceleration is changed. That’s because how you measure acceleration depends on how you measure time. And the Newtonians gave us a standardized time, basically stellar time (time measured in accordance with the rotation of the earth relative to the fixed stars) re-scaled so that on the average a day was 24 hours.

Did Newton’s laws involve arbitrary choices? He could, instead, have proposed Einstein’s relativity which is different from what Newton used. So, clearly, reality did not fix what Newton proposed. But, at the time, Newton’s laws worked very well so satisfied the needs of pragmatism.

The ratchet of pragmatism

If we adopt conventions on a pragmatic basis, and then later we discover a modified set of conventions that would work better, then as pragmatists we should adopt those newer conventions. This seems to be part of how science works. And this gives us a ratchet effect. When we change to something better, then that results in improvement. This ratchet effect is, at least in part, why science seems to always be improving.

But this is not just about science. What separates our way of life from that of stone age people, is mostly the pragmatic social conventions that we follow to organize our way of living. Civilization itself is the accumulated result of that ratchet effect on the social conventions that determine our way of living.

8 Comments to “Conventionalism”

  1. By this logic, words are conventions. Any kind of society or cooperative venture isn’t possible without conventions. That may explain why conventions are so unremarkable. They are truly ubiquitous. Only hermits can claim to oppose conventions, and even they, they can’t vocalize it.

    Now the question is: do convention and conventionalism actually express difference concepts?

    Liked by 1 person

  2. Have you read Helen Longino? I feel like she must have said something relevant to this.

    Liked by 1 person

  3. This applies even to culture or fashion so to speak. At some point it was convention for policemen in the colonies to wear shorts, now everyone is in a long trouser. You even have preachers speaking against women adorning trousers or short skirts. It is only a matter of a time till conventions change.

    Liked by 1 person

  4. Also, Longino is the author of this SEP entry:

    My random thoughts on the meat of your post:
    OK, scientific knowledge happens when some broad consensus of subject matter experts agree that some new theory best accounts for the currently known phenomena of the particular field (I think Longino would agree with this).

    Q: How do we know when someone is a qualified expert?
    A: Roughly, they’ve got a Ph.D. (taught, supervised, and examined by existing experts), and has previously published stuff that the other experts agree isn’t completely off-the-wall.

    Q: What makes this theory the best explanation?
    A: The experts say it is, according to criteria best known to them (because that’s part of what it means to be an expert).

    Maybe that goes some way towards making scientific knowledge a convention among the experts? So, everyone agrees to talk in terms of Aristotle’s metaphysics until Renaissance-era work turns up results that are hard to accommodate, at which point we reach back and revive atomism, add in forces acting at a distance (thanks for that, Isaac), and agree to talk in terms of classical mechanics until the late 19th Century when more puzzling results start showing up, and along comes this Swiss patent clerk, who sets the language for the next 100+ years….
    (I’ve just finished re-reading Alan Chalmers’ _The Scientist’s Atom and the Philosopher’s Stone_, so that bit of history is on my mind).

    Maybe it works. Or maybe that’s taking anti-realism a step too far; I’d have to think about it some more.

    Liked by 1 person

    • A scientific theory is not just a belief system. It is a set of practices (empirical practices). And scientists need to coordinate the way that they follow those practices. That’s why I see it in terms of convention.


%d bloggers like this: