question on grammar literature

Are you talking about this:

[A=>B] => [(not A)=>(not B)]?

My assumptions (as a non-native):
1. "Contingent on" — "Dependent on"
2. "Incidence" — the fact of existance or of having happened (for an event).

I may be wrong though...

EDIT

«... it is generally implied that...»

For a start, can you tell us about your _own_ evidence (an example...) for the use of such an implication? I presume that at least you believe in what you write...

«This is a conditional sentence. IF the contract closing is delayed beyond a certain date, THEN they will turn to another wine.»

OK, I understand. That's just what I wrote in my previous post.

«What I'm trying to argue is that this sentence implies that if the contract closing is NOT delayed beyond that date, then they will NOT turn to another wine.»

Not fully correct. It is not this sentence on its own that makes the implication. There's anoter presumed statement: they'll use their intital wine UNLESS it is impossible.

Now what do we have:

1. D => not(W)
(delay forces them to turn to another wine — it's what the sentence says)

2. not (D) => (W)
(they'll use their intitial wine UNLESS the contract's deadline is violated — the assumtion)

As you see, #2 is doesn't directly follow from #1. It is just a given assumption, which is evident for everybody and thus not explicitly declared.

Here's what happens without such an assumtion:

IF an object is a stone THEN it is rigid.

But it in NO WAY implies that
IF an object is NOT a stone THEN it is NOT rigid.

P.S.: Sometimes I am weak at making good examples, though I am sure you'll find better ones in many books on logic. You should look for: a fallacy called "Argumentation from ignorance" (it should also have a Latin name).

My advice would be not to forget about the assumption in your dissertation, because its presence makes the implication "logically coercive" (using your terms), while the lack of it renders it downright incorrect.

EDIT:
As to the the degree of contingence that you mentioned ("equally contingent", you said), I don't think it is correct under any self-evident assumtions.

1. The wether being good, we'll certainly go for a walk (95% contingence)
2. The wether being rainy, we'll probably make a stroll anyway (60%).

You see, there's no reason to consider the contigence of not going for a walk (100-60=40%) under the rain equal to that of making a stroll (said 95%) under the sun.

Hope you understand me.

«But that is precisely why I said that it's not "logically coercive" and only "generally implied" -- to make it clear that not every conditional sentence can be inverted in this manner.»

OK. But it is not "generally implied". I'd rather say it is in many cases correct to derive an event's non-incidence from that condition's having not been met.

It happens due to ambiguous meaning of "if" in natural language. In everyday speech sometimes "if" means "when and only when".

— Will you go for a walk tomorrow?
— Yes, if the wether's good.

The answer is equal to: "I'll go only if ((if and only if) the wether is good".

«Again, this is why I am looking for literature on the subject -- to find out if there is some universal distinguishing element between conditional sentences from which such an implication can be drawn, and ones from which it can not»

In addition to what I said above, I can direct you to what linguists call presuppositions. That's how they're a defined in Wikipedia:

«In linguistics, a presupposition is background belief, relating to an utterance, that:
1. must be mutually known or assumed by the speaker and addressee for the utterance to be considered appropriate in context
2. will generally remain a necessary assumption whether the utterance is placed in the form of an assertion, denial, or question, and
3. can be associated with a specific lexical item or grammatical feature (presupposition trigger) in the utterance.»

I think it isn't easy to arrive at some formal way to distinguish such things because it's a matter of our way of thinking, a problem of Natural Language Processing, because the distinguishing whether an "if" means "if" or "if and only if" would require certain knowledge about the universe. Furthermore, this problem seems language-dependent...

So, if such a "universal distinguishing element" (as you call it) even exists, it's the presupposition.

«I am going to check on "argumentation from ignorance", but I am afraid that this term will mostly be used to express the notion that people are generally being ignorant in their argumentation»

Not quite right. This type of logical fallacy is a particular case of deriving (not A) => (not B) from A => B.