Romance: Philological curiosity

[Guijarro]

Some of you, people, might be philologists on top of being linguists, and you will be able to explain this curious fact.

I have always been at a loss to figure out why a romance word, SALIR, has ended having three altogether different meanings in three modern romance languages that I sort of speak and understand.

In French, SALIR, means to dirty
In Spanish, SALIR means to go out
In Italian, SALIR(e) means to go up

How can these absolutely different meanings have evolved from the same (or is it?) Latin root?

[freknu]

French "salir" appears to come from Proto-Germanic, while Spanish and Italian "salir" both come from Latin "saliō".

PG. *salwaz "dusky, dark; dirty, muddy"
Old Frankish, *salo "dull; dirty; grey"
Old French, sale "dull; dirty"

Lat. saliō "to leap, spring, spring forth"

[Daniel]

[BTW, I just added a prefix of 'Romance:' to your title to get things going in this forum]

Interesting. I hadn't noticed this. Of course I know the word in Spanish, but I don't know if it in French and can't remember if we talked about it in my Italian classes.

I imagine the Spanish and Italian are different, with at some point the direction changing ("go up out of the box", etc.).
No clue about French.

"go out" is a little contentful, but it's still somewhat general like "go", and think about that in English:
go out
go up
go around
go "Hello!" (He went "hello!")
go crazy
??go dirty
go out of your way and get dirty

It's not unimaginable, but it is interesting.


To extend the list a bit:
Catalan: to leave / go out
French: dirty / soil [=entered English too, I think?]
Galician: sair 'go out/leave'
Italian: salire 'go up', 'ascend', 'advance'
Latin: salire 'jump', 'dance(?)'
Portuguese: 'leave', 'exit', 'go out'
Spanish: 'go out'/'leave', 'emerge', 'rise'
Romanian: sări [cognate, I think] looks like Latin: 'jump', 'dance'

Source: Google Translate [as a starting point]



Observations:
1. Iberian Romance clearly has the 'Spanish' meaning (maybe borrowed from Castilian, maybe older Iberian Latin/Romance).
2. It is possible to imagine a semantic shift from 'jump' > 'go up' > 'go out', so I think Latin/Italian/Spanish are related, with Romanian preserving the original Latin meaning.
3. I think Western Romance (broadly, including Italian [Florentinian?]) changed to 'go up' from 'jump'.
4. I think then Iberian changed to 'go out', maybe also French (so Western Romance narrowly).
5. If the French form is related (I'm not sure), I would (just) guess that it came from something along the lines of "oh, look, the kids went out to play and their clothes are dirty!" or "let's fight a battle and get dirty!"-- a resultative sense, where going outside comes with dirt.


And I wonder (just wonder) whether somehow the meaning of "go out" might be associated with the leaving/migration/battles involved in going westward for the Iberian Romance languages. It could explain why they evolved a new form, if for some reason they used it a lot-- "jump to your feet soldiers, let's win this battle!" and change followed...


---
Freknu (posted while I was typing this):
That makes sense. So French is the exception. Ok

[Guijarro]

I knew you would have good explanations for this difference in meaning. Thanks to both of you!
Too bad (for my subsequent musings!) that Freknu has definitely established that the French word comes from another root.

Well, tough!, I'll have to muse with the only two left that come from the same one.

Daniel has given us a beautiful story about how they might have changed (the three of them! Pity we have to forget about the French, now). It all sounds possible and very intelligent...

...FOR US, HUMAN BEINGS!

Imagine, though, that we are a species with a fixed code system we use in our communications. Can you just imagine that situation?

Now, if you have, how would you be able to follow Daniel's story and understand it? If our ONLY way to communicate is using coded material, the history of the three (now, just two) semantic versions of the same original term would have been almost impossible for they are fixed by our code. We never visualize this difficulty, of course, since our way of communicating is chiefly inferential and we are able to establish pragmatic meaning relations with different conceptual elements that have some (inferred) similarity for us. We do that so easily that we never ask how this is indeed possible. It is not only possible, it is obvious!

Well, it isn't!

Do we want a more clear evidence to show how human communication is CHIEFLY inferential? I can't think of a better one!

[Daniel]

I'm glad my explanation makes some sense (although it's a little worrying that the explanation for French also 'made sense' even though freknu showed us it was wrong!).

As for French, is sortir possibly etymogically related? R/L sound changes occur, but I can't quite imagine that specific one within Romance.
[Maybe it was in Latin-- the dissimilation of r-r only occuring when not interrupted by another consonant? And it's from the perfective stem, like English -ate endings? So French represents an older phonological form in Latin (R) within a secondary/more specific meaning of the word. I don't know if any of this is right, though.]


As for the rest

....
Do we want a more clear evidence to show how human communication is CHIEFLY inferential? I can't think of a better one!

What I believe this shows is that communication is not perfect: the encoded message and the decoded message are not identical-- in a technical sense, language is not a code or an "encryption" system. I don't think that it directly proves your point.
However, given then that language is not a code, I think that necessitates pragmatic inference-- otherwise we'd just be approximating communication without any principles beyond imperfect signals.
So, yes, with an extra step in the middle, I think it does support your claim.

Imagine, though, that we are a species with a fixed code system we use in our communications. Can you just imagine that situation?

Fixed at what level? UG? Or surface form (including lexical items)? Or something else?

[Guijarro]

What I believe this shows is that communication is not perfect: the encoded message and the decoded message are not identical-- in a technical sense, language is not a code or an "encryption" system

Communication is not perfect from the coding/decoding point of view, I agree; however, if we forget about the coding/decoding prejudice, we will have to admit that human communication works beautifully for it uses our inferencing powers AND the advantages of the current code which, as is shown in the new branch of Pragmatics (i.e., "Lexical Pragmatics"), can however be adjusted to new situations with total ease. We have, thus, three advantages on other species' communicative acts:

(a) To begin with, we are able make out what the intentions of the speaker are without recurring to a previous code which would restrict the possibility of our communicative desires. We do that so well, that Grice called it "mind reading", if you please.

(b) But then, on top of that, we are able to eat the cake and keep it, for we use also a set code to help us in accuracy. So we may distinguish, for instance, between our mind state in situations where, say,

e.1. I am 100% certain about P
e.2. I am not 100% certain about P

because our code distinguishes between those two states (i.e., KNOW and BELIEVE)

(c) We may further CHANGE the code if by some reason it doesn't fit exactly to our communicative desires. So, as the code is not fixed, we may make it more and more accurate, thereby creating new terms, which may not be eternal, and may change again when needed.

This last advantage, however, is responsible for the Tower of Babel's mess.

But that's how life is: imperfect

[Daniel]

To be clear, I'm just talking about a technical definition of code, which (I believe) is one in which a message can be encoded and then decoded in a 1-1 relationship. Clearly that's not the case with language-- pragmatics is that additional component.


Otherwise what you said basically makes sense to me.