everyone, I'm reading Kees Doets's Basic Model Theory (which is freely and legally downloadable from https://web.stanford.edu/group/cslipublications/cslipublications/Online/doets-basic-model-theory.pdf). There are two different definitions of logical consequence in it. The official one ($\vDash$) is on p.6:

$\Sigma$ is a set of formulas, the notation $\Sigma \vDash \phi$ -- $\phi$ follows logically from $\Sigma$ -- is used in the case that $\phi$ is satisfied by an assignment in a model whenever all formulas of $\Sigma$ are.

The other, 'unofficial' one ($\vDash ^*$), is in Exercise 8, which is on p. 7:

Sometimes, logical consequence is defined by: $\Sigma \vDash ^* \phi$ iff $\phi$ is true in every model of $\Sigma$.

It bothers me how these two definitions are different. Indeed, this is exactly what Exercise 8 asks:

Show that if $\Sigma \vDash \phi$, then $\Sigma \vDash ^* \phi$, and give an example showing that the converse implication can fail. Show that if all elements of $\Sigma$ are sentences, then $\Sigma \vDash \phi$ iff $\Sigma \vDash ^* \phi$.

Let me say what I think about the first definition of logical consequence, i.e. $\vDash$. It seems to me that

$\phi$ is satisfied by an assignment in a model whenever all formulas of $\Sigma$ are.

just means

whenever all formulas of $\Sigma$ are satisfied by an assignment in a model, $\phi$ is satisfied by the same assignment in the same model.

which, appears to mean

for all models $\mathcal{A}$, for all assignments $\alpha$, for all $\psi \in \Sigma$, if $\mathcal{A} \vDash \psi [\alpha]$, then $\mathcal{A} \vDash \phi [\alpha]$.

I think I'm right about $\vDash$. But I'm less certain about $\vDash ^*$. From p. 6 of the book,'$\mathcal{A}$ is a model of $\phi$' is just:

$\mathcal{A} \vDash \phi$

which means

$\phi$ is satisfied in $\mathcal{A}$ by every assignment.

Therefore, the right hand side of the definition of $\vDash^*$

$\Sigma \vDash ^* \phi$ iff $\phi$ is true in every model of $\Sigma$.

is just

For all $\mathcal{A}$, for all $\psi \in \Sigma$, if $\mathcal{A}\vDash\psi$, then $\mathcal{A} \vDash \phi$.

However, if what $\mathcal{A} \vDash \phi$ means is just that $\phi$ is satisfied in $\mathcal{A}$ by every assignment, then (and perhaps it is where I have made crucial mistake which I don't understand) I don't see why it cannot be translated also as

For all $\mathcal{A}$, for all assignments $\alpha$, for all $\psi \in \Sigma$, if $\mathcal{A}\vDash\psi [\alpha]$, then $\mathcal{A} \vDash \phi [\alpha]$.

which is just identical to the definition of $\vDash$!

This elementary problem bothers me quite a bit and I hope someone can tell what mistake I have made. Many thanks in advance.


Your analysis of $\vDash$ is fine, but you've shuffled some (important) parts of the definition of $\vDash^*$. The latter definition says: For every $\mathcal A$, if all assignments in $\mathcal A$ make all the formulas in $\Sigma$ true then all assignments in $\mathcal A$ make $\phi$ true.

The crucial difference is that here the "all assignments" quantifier is applied separately to the "make all of $\Sigma$ true" assumption and the "make $\phi$ true" conclusion, whereas in $\vDash$ the "all assignments" quantifier is applied to the whole implication "if it makes $\Sigma$ true then it makes $\phi$ true."

The underlying general fact here is that quantifiers $\forall x$ cannot be distributed across implications. $\forall x\,(P(x)\to Q(x))$ is not equivalent to $(\forall x\,P(x))\to(\forall x\,Q(x))$. Example: It's true that "if all people are American then all people are left-handed" because the antecedent and consequent are both false. But it's not true that "all Americans are left-handed."

  • $\begingroup$ Thanks! I think this is what I want, i.e. the difference in the logical forms of the definitions. $\endgroup$ Nov 27 '21 at 2:20

I think it's best to take the question's hint and understand that the two notions are not equivalent by looking at a counter-example.

Let $\Sigma$ contain only the formula $\forall x (x + y = x)$, and let $A$ be a model of this formula: i.e., for every assignment of a member of $A$'s domain to the variable $y$, the formula is true. This is essentially to say that $x + y = x$ holds for all $x$ and for all $y$ in $A$, because we can assign any member of the domain to $y$. So it turns out that $A \models \forall y \forall x (x + y = x)$. This establishes:

$$ \Sigma \models ^{*} \forall y \forall x (x + y = x) $$

Is it then also the case that

$$ \Sigma \models \forall y \forall x (x + y = x) $$

i.e. for any model $B$ and an assignment to $y$ from the domain of $B$ which satisfies $\forall x (x + y = x)$, does $B$ also satisfy $\forall y \forall x (x + y = x)$? Try to construct a model and assignment where this isn't true. Hint: think of the natural numbers and assign $0$ to $y$.

The subtle point is that, in reasoning about $\models ^{*}$, we require that a model $A$ satisfies all of $\Sigma$ on any assignment in $A$. With plain old $\models$, we only require of a model $B$ that a particular assignment in $B$ satisfies $\Sigma$.

  • 1
    $\begingroup$ But why then the converse is always true if ⊨ is a lot "looser" than ⊨∗ with open wff(s) in Σ for both cases? $\endgroup$
    – mohottnad
    Nov 25 '21 at 7:55
  • $\begingroup$ @mohottnad Yep, that was poor wording on my part, which I've removed. $\endgroup$ Nov 25 '21 at 12:23

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.