Horst Nowacki: Semantic Precision and Logic in STEP

Horst Nowacki presented some comments in response to Nicola Guarino's talk raising issues of deficient semantic precision in STEP. The question addressed is mainly: What is semantic precision and how do we achieve it?

The presentation was mainly based on two quotations from famous authors (See Encl. 1 and 2):

• What do words mean?
  Brief quotation from Lewis Carroll, Through the Looking Glass (1872), famous dialogue between Alice and Humpty Dumpty.
• Miscellaneous excerpts from Ludwig Wittgenstein's Tractatus Logico-Philosophicus (1921) dealing with the meaning of names and sentences.

These quotations speak mainly for themselves and shall not be explained here. But some conclusions will be attempted regarding the achievable semantic precision in information models, in particular in STEP.

What does Lewis Carroll tell us in this regard?

• Words do not have an absolute meaning by themselves.
• Words are given an intended meaning by the person who uses them.
• Words are symbols to designate intended meaning.
• Fortunately only the intended meaning matters in communication.
• The intended meaning can be explained in sentences by means of other words.
• Common usage of words is not a precise yardstick for intended meaning.

My conclusion for STEP:

The criticism that STEP entities (like words) are semantically imprecise and do not mean what they ought to mean seems beside the point. The issue of what an entity should precisely mean cannot be settled at the entity (or word) level alone.

This is where Wittgenstein comes in and helps. The key point for my purpose is his paragraph 3.3: THE SENTENCE HAS MEANING; a name has meaning only in the context of a sentence.''

My conclusion for STEP:

Names correspond to objects at entity level, sentences may be compared to EXPRESS schemas. The Application Reference Model (ARM) in particular consists of one or several schemas. This defines a domain of discourse (like a sentence context) and claims to capture the requirements of an application. It is legitimate to criticise lack of semantic precision at this level. It is possible to test the model for sufficiency in meeting the application requirements.

Valid criticisms of STEP models are possible, in particular:

• if the verbal description of an entity in some application context does not agree with the meaning formally documented in the EXPRESS model (e.g., missing constraints, incomplete set of properties),
• if the EXPRESS model does not agree with the intended meaning so that a domain of discourse cannot be adequately covered.

Even if a STEP model at ARM level is not semantically precise for the full domain of discourse which it claims, it may be free of contradictions and precise enough for a reduced domain. Its limitations should then be clearly stated.

Summary:

There is room for improvement in STEP with regard to semantic precision at the ARM or application level where the requirements of the model are stated. It may also be possible to bring verbal entity descriptions into closer agreement with their intended meaning in the context of a schema. But it seems impossible to judge semantic precision at the entity level alone. Therefore I do not understand some of the objections raised against lack of rigour in STEP.

Enclosure 1

What do words mean ?

``I don't know what you mean by 'glory' '', Alice said.
Humpty Dumpty smiled contemptuously.

``Of course you don't - till I tell you. I meant 'there's a nice knock-down argument for you' !''

``But 'glory' doesn't mean 'a nice knock-down argument' '' , Alice objected.

``When *I* use a word'', Humpty Dumpty said, in a rather scornful tone, ``it means just what I choose it to mean - neither more nor less.''

``The question is'', said Alice ``whether you *can* make words mean so many different things.''

``The question is'', said Humpty Dumpty, ``which is to be master - that's all''

Lewis Carroll, Through the Looking Glass (1872)

Enclosure 2
Quotations from L. Wittgenstein,
``Tractatus Logico - Philosophicus''

3.202

The simple tokens used in a sentence are called names.

3.203

The name means the object, the object is the meaning of the name.

3.22

The name substitutes for the object in the sentence.

3.26

The name cannot be further subdivided by any definition: It is a primitive token.

3.263

The names of primitive tokens can be explained by comments. Comments are sentences that contain the primitive tokens. They can thus be understood only if the meaning of those tokens is already understood.

3.3

Only the sentence has meaning; a name has meaning only in the context of a sentence.

3.32

The token is that which is sensually perceivable of a symbol.

4.021

The sentence is an image of reality.

4.026

The meaning of the simplest tokens (the words) must be explained to us so that we can understand them. But we communicate by sentences.

ProKSI-97 Report