This question is partially related to How does "is" work? and What is the difference between the "is" of predication and the "is" of identity?, but more specifically it concerns statements about physical quantities and measurements.

[Question heavily re-edited because unclear]

Consider the statement "the instantaneous velocity of body B is 1 m/s". This is usually expressed in a mathematical form like "v = 1 m/s" or similar.

The "=" in that form seems to suggest that this statement expresses a relation. But is this really the case? or does that statement express a predicate?

The question is also related to the fact that we can consider the statements

"the instantaneous velocity of body B is 1 m/s"


"the mean velocity of body A is 1 m/s"

and deduce

"the instantaneous velocity of body B is the same as the mean velocity of body A

All three statements – which are perfectly fine – can be expressed in deceivingly similar mathematical forms: "v = 1 m/s", "u = 1 m/s", and "v = u".

The last expression is particularly nasty, because "=" or "is the same as" really means "has the same value as", and not "is the same (thing) as". The latter would obviously be false, because mean and instantaneous velocity are different notions, and moreover they refer to different bodies.

By the same token, "the instantaneous velocity of body B is 1 m/s" seems to really mean "the instantaneous velocity of body B has the value 1 m/s".

Clearly(?) the statement "v = u" expresses a relation ("has the same value as"), and the relation is symmetric, reflexive, and transitive; an equivalence relation.

But does "v = 1 m/s" also express a relation? It seems to me that it doesn't. The chain of reasoning {"v = 1 m/s", "u = 1 m/s", "v = u"} seems to use transitivity and symmetry, and suggests that "=" is the same along the chain, and it's therefore a relation.

And yet I think that "=" denotes different things in the chain. It's mixing apples and oranges. And yet, graphically at least, it seems to work.

My questions are:

  • Does the statement "v = 1 m/s" (or "the instantaneous velocity of body B has the value 1 m/s") express a predicate? Or a relation? Or something else? Or is the intepretation arbitrary? (I would be inclined to see it as a predicate, but in a second-order logic with an infinity of such predicates, over which we can use quantifiers.)

  • How does an inference chain such as {"v = 1 m/s", "u = 1 m/s", "v = u"} work, if the "=" has different meanings in it? (eg if it expresses a predicate in the first two statements and a relation in the last.)

  • Can you suggest any literature on this issue, in relation to measurement statements in science?

I can add that a similar question arises in probability theory, which typically has statements about "random variables" of the form "X = x". This would mean "the quantity X has value x". But such an interpretation has lead to a great debate about causality and correlation. The problem is that more precisely "X = x" can mean two things: "X has been observed to have the value x" and "X has been set to the value x". These two very different statements have hugely different consequences, for example if they appear in the conditional of the probability of another statement.

[Note: I'm a physicist and this question is very important for my field, so I'd like to keep some science- or physics-related tag for this question.]

  • 1
    When we say that "the velocity of body B is 300 km/s" we are asserting that the value of the magnitude velocity is ... Jun 2 '20 at 7:53
  • 2
    When we write the mathematical formula "v = 300 km/s" we are asserting the same fact: the value of the "function" velocity (space/time) for a given body. Jun 2 '20 at 7:54
  • 1
    In both cases, "is" is not apredicate. Jun 2 '20 at 7:55
  • 1
    Why is my original "physics" tag inappropriate?
    – pglpm
    Jun 2 '20 at 8:01
  • 1
    In physics, there is no doubt as to the fact that it is an " is" of identity. But there is actually a metaphysical issue which is adressed is Aristotle's categories. You may have a look at the SEP entry on " Aristotle's Categories" and on " Aristotle's Metaphysics" ; also, Marc Cohen's personal website. For Aristotle , " Peter is 6 feet tall" attributes the predicate " 6 feet tall" to Peter.
    – user37859
    Jun 2 '20 at 11:05

The statement "The ball is red" can be rewritten with subject-predicate form: "Red(ball)" where "Red( )" is a predicate (a property predicated of something) and "ball" is the subject (an object of which the "redness" is predicated).

In this form, there is no "is". This is the background for the assertion that, in statements like that above, "is" is not a predicate.

But "is" means also equality: a logical binary relation.

We use it ubiquitously in mathematics and physiscs: for example in equations, like e.g. 1+1=2 and F=ma (see e.g. the post What exactly is an equation?).

When we say that "the velocity of body B is 300 km/s" we are asserting that the magnitude velocity has a certain value.

Thus, when we write the mathematical formula "v = 300 km/s" we are asserting the same fact: the value of the "function" velocity (space/time) for a given body.

This means that, IMO; "v = 300 km/s" is not an equation (an identity between e.g. numbers or functions).

The formula "v = 300 km/s" is like an assignment in logical programming: "let the value of v be..." that we use in describing a physical problem, or the result of an operation of measurement.

Some useful references:


  • I was wondering if, when you are saying "a property predicated of something", you really want to make an ontological commitment or whether this is just an accident. I mean, one can -- as I did in my answer -- perfectly well say "Red()" is a predicate without committing to properties. Thx. for a clarification.
    – Mr. White
    Jun 2 '20 at 14:46
  • Maybe your explanation can be put as follows: we have a function F meaning "has the value", so we have "F(v)=1m/s" and "F(u)=1m/s", and we can also say "F(v)=F(u)". What confuses me is that "F(v)=1m/s" seems to be a meta-statement, rather that a statement within the formal system, whereas "F(v)=F(u)" is within the system. Am I wrong? And would the deduction "F(v)=1m/s", "F(u)=1m/s": "F(v)=F(u)" be within the system, or on a meta-level? And also, what are the objects in the range of F?
    – pglpm
    Jun 3 '20 at 10:23
  • 1
    @pglpm - as per my answer, the issue is a little bit thorny... We manage physical equations in the same way we manage "usual" equation: if we have the expression (x-1) and we let x=1, we substitute the value of x into the equation and we compute the result: (x-1)[x ← 1]=0. The same for a physical case: Ek= 1/2 m v^2, for a body with mass 2 and v=2 we substitute to compute the kinetic energy: Ek=v^2=4. The computations are performed in the same way using the properties of equality: Jun 3 '20 at 10:52
  • x = y → (P(x) → P(y)) where P( ) is a formula. My issue in treating the formula v=2 as a "standard" equation is that from m=2 and v=2 we cannot meaningfully conclude that m=v, i.e. that the magnitude mass and the magnitude velocity are identical. We have to be consistent and asserting that "the value of the magnitude mass and the value of the magnitude velocity are (in that specific context) identical". Jun 3 '20 at 10:55

This is less a question of how "is" works and more a matter of what exists in your ontological realm.

If "the velocity of body B" refers to an object and if "300 km/h" refers to an object, then "The velocity of body B is 300 km/h" expresses the identity of these two objects.

Let's, for the sake of the argument, say "red" does not refer to an object in your realm. Then "The sun is red" does rather not express an identity and "is" is rather used predicatively. Note that this does not imply the existance of qualities (or alike).

Recommendable literature: "On what there is" and "The Philosophy of Logic", both by W.V.O. Quine.

  • Thank you for these thoughts. The question is also how "300 km/h", per se, can refer to an object, and what does this mean. We could of course introduce a set of "individuals" and a set of "qualities" and introduce the relation "... has the quality ..." between the two sets. But this seems a roundabout way to introduce predicates.
    – pglpm
    Jun 2 '20 at 8:27
  • To say that "is" is used predicatively, as I did above, does not ontologically imply that there exist qualities -- I have added that to my answer. This is a decision an ontologist has to make. However, I am very surprised that a physicist does go into such philosophical depth. Yet, I don't know much how physicists work.
    – Mr. White
    Jun 2 '20 at 9:34
  • I understand that it's up to us how to model such a statement. That is, I could consider it within propositional logic and avoid speaking of predicates and functions; or within some first-order logic considering it as a predicate among many or as a relation or as a function; or within some second-order logic so that I can take "300 km/s" an an argument and quantify over it...
    – pglpm
    Jun 2 '20 at 10:33
  • The problem is that some mathematical or physical statements do not make sense depending on the intepretation. From the statements "v1 = 300 km/s" and "v2 = 300 km/s", where v1 and v2 refer to velocities in completely different physical contexts (say, an instantaneous velocity vs a mean velocity), I could deduce "v1 = v2", which from some interpretation does not make sense: an instantaneous velocity is not a mean velocity – they have different definitions.
    – pglpm
    Jun 2 '20 at 10:40
  • 1
    I would agree that the sheer use of "=" does not guarantee identity. "=" could be used contextually and in a certain context not mean identity.
    – Mr. White
    Jun 2 '20 at 11:40

A few thoughts:

-do all of these conceptual questions still cause difficulties if everything is expressed purely in mathematical terms and in full without abbreviations? I don't have the right typeset on my computer but, for example, I think that if you typed out the formula for mean velocity in full it would be quite clear what is meant by saying that it some case a given mean velocity is equal to a given instantaneous velocity.

-talking about whether motion is a property or a relation seems to me to be closely related to the general discussion about substantivalism vs relationism in respect of space and time; if you haven't already read it John Earman's World Enough and Spacetime is a great book on this

-re the notion of equality vs equivalence in mathmatics; I believe that there is a whole area of research on this at present, with Jacob Lurie at IAS playing a leading role. See this popular article at Quanta https://www.quantamagazine.org/with-category-theory-mathematics-escapes-from-equality-20191010/ - This sounds fascinating, but Lurie's 944 page book 'Higher Topos Theory', however, sounds a very long way further into the domain of hardcore pure mathematics than I am ever likely to venture!


The question seems to have both a logical aspect and an ontological one.

But first , let me try to explain how I see the reasoning that is behind the question.

(1) You are not willing to accept that instantaneous velocity and mean velocity are the same thing ( with sameness = identity). (In particular , you are not willing to admit that the instantaneous velocity of an object A be the same thing as the mean velocity of an object B).

(2) But you encounter this perfectly possible case : v = 1 m/s and u = 1 m/s with v denoting the instantaneous velocity of object A and u denoting the mean velocity of object B

(3) And here, a problem arises, because identity is symmetric transtitive. So , if you admit (2) , you are forced to say that : v = 1 m/s and 1 m/s = u , in such a way that : v = u

(4) Since the final step of (3) is apparently incompatible with your principle (1) , you want to escape this situation in the following way :

in (2) the assertions should not be expressed as equalities, but as subject - predicates sentences. The subject - predicate relation is not symmetric and transitive, so in that case, the undesired result at (3) is avoided.

What I would like to show is that there si no contradiction betweeen (1) and (2)-(3) .


  • You can first say that you do not identity a physical reality ( a force, a quantity) with its measure, its value. ( after all, a measure is only a number, but physics is not ultimately about numbers, abstract entities ; it is about concrete znd causally efficient reality)

  • But your symbols " v " and " u" denote the values of the physical realities " instantaneous velocity" and " mean velocity" . ( Meaning : the two symbols do not denote the physical realities themselves).

  • So, saying that v = 1 m/s ( identity) and that u = 1 m/s ( identity) , implying that v = u ( identity) is perfectly OK and harmless.

  • The two velocities ( physical realities) are simply equivalent in values ( as you pointed out) , while their values are really and unproblematically identical ( one and the same number).


Note : here , one can make use of a distinction that we owe to Frege between the denotation of a symbol and its sense ( the concept that it expresses).

  • For example : " the morning star" and " the evening star" denote the same object, but do not express the same concept.

  • The statement "v = u" becomes harmless when we consider it in a purely extensional way. In its extensional reading ( the one that mathematical science favours) it simply means

the number denoted by v is the same number as the number denoted by u

  • But that does not make your principle (1) false ; because your principle is still valuable at the intensional level :

in spite of the identity v = u , there is no identity between the concept of v ( that is, the concept of instantaneous velocity ) and the concept of u ( the concept of mean velocity)

  • 1
    Thank you for this point of view. I'm not sure if we can really conflate "having/assuming a numerical value" with "denoting a numerical value". Say, "age(Bob) = 20 years". Does the age of Bob denote 20 years? or refer to 20 years? I'm not even sure I would understand what that means...
    – pglpm
    Jun 2 '20 at 13:56
  • 1
    It seems my question deals with something that hasn't been settled yet.
    – pglpm
    Jun 2 '20 at 13:57
  • 1
    you're right, the question is not an easy one
    – user37859
    Jun 2 '20 at 14:01
  • 1
    I appreciate all points of view and literature proposed so far nevertheless
    – pglpm
    Jun 2 '20 at 14:01
  • 1
    wow greatly extended answer!
    – pglpm
    Jun 2 '20 at 17:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.