Re: [ontolog-forum] Ontologies and languages

[William Frank <williamf.frank@xxxxxxxxx>]

On Thu, Jun 18, 2015 at 2:59 PM, Thomas Johnston <tmj44p@xxxxxxx> wrote:

William:

Very well put. 

I would suggest, however, that pragmatics rests as much on Grice as it does on Austin. 

Thanks, Thomas, I did not realize that the concept of conversational implicature came from Grice, till i just looked it up, or that their ideas were developed at the same time, only Austin published more earlier.   I now remember reading 'logic and conversation"

In several places in my book Bitemporal Data: Theory and Practice, I attempt to extend Grice's rules of conversational implicature to the statements that we make by inserting, updating and deleting rows in relational databases.

cool.
 

For example, by inserting a row in a table, the party performing that action intends, unless he otherwise indicates, to express his belief that the row makes a true statement, and intends those retrieving the row into a query result set to understand that that was his intention. And those of us who look at rows in our result sets do indeed understand that those rows were created in accordance with those implicature rules.(I also describe Gricean rules for updates and deletes.)

I also attempt to extend Austin's speech acts to those speech acts which express propositional attitudes, in particular to the speech acts of asserting that a statement is true, and of withdrawing such an assertion --

I did something similar in an abandoned patent from the late  1990s, both looking at the various implicit speech acts in different computer communications, eg. some files of transactions are requests that the transactions be executed, others are assertions that the transactions have occurred, and proposed an XML form which was always a triple of an actor, a speech, a proposition or open sentence, and a speech act modifier.   

thus defining a third temporal dimension in which every row in a relational table is related to every person or group who might agree that the row is true, or disagree. For each pair <row, person/group>, there is an assertion time period which begins when that person/group asserted that the row made a true statement, or assented to such an assertion, and ended when that person/group withdrew that assertion or assent.

This is very nice, and very important, in my opinion, in that it directly links 'what the computer says' to some responsible party.   

Tom Johnston

On Thursday, June 18, 2015 1:56 PM, William Frank <williamf.frank@xxxxxxxxx> wrote:

Hello, Juan

Let me try to answer your question about domain ontologies and languages, from the point of view of mathematical logic and linquistics, leaving philosophical ontology to the side.   I think what I say is consistent with what Thomas Johnson and Ed Barkmayer said, but I define more terms.

First, when one talks about language in this arena, let us consider these languages to be formal languages.  That is, a formally defined language, rather than a natural language, even it looks like special a bit of English or Chinese.

Logicians and mathematical linquists have carved up formal languages more or less thusly.

1. a syntax that specifies what configuarations (usually strings) of symbols are grammatically correct sentences in the language.   This says nothing about what those sentences might 'mean'.   Often, when logicians say 'language', they just mean some specified set of uninterpreted strings. 

2. a 'logic', comprising

2a. A deductive system that provides production rules as to what expressions can be generated from what others, intended to preserve one or more logical relations between the initial and generated expressions (such as, in order of the strength of the relation,  tautology preserving, consequence preserving, or congency preserving).

2b. a sematics that specifies what the consequence relation, and other related relations.   This give us the 'meanings of the sentences in a very narrow sense.    The sense it which the fact that a sentence A is true implies that any sentence of the form (A or B) is also true.  (From which it follows that if both A and B are true, so is (A or B).    This gives us  not only logical consequences but also tautologies or 'logical truths'.  

3. Theories --

3.a.  An a-priori theory is a set of sentences taken to be definitions and axioms that describes meaning relations among terms in the language, using some of the sentences in the language.   For example,  the assertions that less than is transitive, that zero has no predecessor, that 0 plus any number is that same number.    The theory tells you not only the meaning of logical structures, but lets you understand what the 'words' mean.     So, while 'No square has only three sides' is not a truth of logic, it is true by definition.  You discover it by knowing what the word 'square' means and knowing how to reason, but by examining ever more squares.

3b. An empirical theory  is a set of sentences that describe observable facts.  Water boils at sea level at 100 degrees centigrade.    The Egyptian pyramids have square bases.

4. The pragmatics of the language, which provides a level of rules for honest communications that go beyond those of the logic.  This is based today on the theory of speech acts.   For example, though it might be *true" that A or B is true, whenever A is true, you mislead your audience if you tell them "A or B" when in fact YOU know that it is only A that is true, and you know that B is false or you know that B is also true.  This is based on something called the principle of maximal information -- you have a social contract to covey as much as you can in the space provided, not say less using more space, so that there is a conversational implicature that when you say A or B, you do not know exactly A only.  (and, before the principle of maximal information and the study of pragmatics, when there was only logic, people posited that there were two 'meanings' for the english word 'or' - inclusive and exclusive.)  Instead, there is only one logical meaning, and a much richer conversational implicature.   For example, if the flight attendant said "we have coffee or tea', and you asked for coffee, the flight attendant could not say, 'sorry, we have only tea!" even though the exclusive or (exactly one) would permit that usage.)   

5. The personal conntations of certain sentences and phrases, that vary greatly depending on one's experience, but are still widely shared (otherwise, there would be no aclaimmed poetry).    For example, someone on this forum said "In the 1030s, the word 'war' meant the first world war."   This is yet another meaning of the word 'meaning." 

People tend to jumble all these together, and every natural language has all these aspects.   But if we keep them separate, then we can answer questions such as the ones you posed more clearly.

A domain ontology is an a-priori theory about the meaning of terms in a domain.    You can't therefore have a domain ontology, feature 3a.  without having features 1. 2a and 2b.  If you mean that 1, 2, and 3a is what you have to have to have a 'language', why then, an ontology is a language.  

And, this is what a language might mean to a chemist.  Roughly, the language of chemistry would tell us that hydrogen has one electron and oxygen seventeen, and what exothermic means. The emperical science of chemistry, on the other hand, using that language, might say that says that hydrogen and oxygen combine exothermically.   

(Of course it is not so simple, in that the 'languages' of sciences depend themselves on deep discovers, such as the periodic table.   And because ontologies describe, to some degree, the behavior of a language community, so are related to emperical observations of a kind.  OTOH, without the meanings of the words being fixed, we can't communicate about the facts.   If you say tomato, but you mean potato, the world does not go round.)

So, the question 'is an ontology a language?' by itself, is not sufficinetly clear. 

Similarly, an ontology surely does not DEFINE a language, in the sense that the syntax comes before any a priori theories.

   

On Wed, Jun 17, 2015 at 9:59 AM, Juan de Nadie <juandenavas@xxxxxxxxx> wrote:

Hello all,

After reading several papers about ontology, I think that I'm not completely sure about the relationships between ontologies and languages.

When we build a domain ontology, its set of categories and relations can be used for developing a domain language for representing specific state of affairs of tis domain. With this in mind, as an ontologist, I was trying to understand the relationship between the ontology and the language developed using its set of categories and relations.

• An ontology is a language? I found some papers that say something in this way. In this case, it seems that we are collapsing the notions of ontology and language or, at least, we are assuming that there is a subsumption between them.
  
• An ontology defines a language? In this case, ontology and language are considered different types of entities. However, in this case, we can ask what are the differences between them.
  
• Is there another option?

Thank you in advance.

Best regards.

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
 

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
 

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (01)