add document on semantics options and add options to terminology document

docs/rdf-terminology.md

@@ -134,3 +134,11 @@
 * [Valid](https://w3c.github.io/rdf-semantics/spec/#dfn-valid) — Any process which constructs a graph E from some other graph S is (simply) valid if S [simply entails](https://w3c.github.io/rdf-semantics/spec/#dfn-simply-entail) E in every case, otherwise invalid.
 * [Value space](https://w3c.github.io/rdf-semantics/spec/#dfn-value-space) — The value space of a datatype is the range of the [lexical-to-value mapping](https://w3c.github.io/rdf-concepts/spec/#dfn-lexical-to-value-mapping).
 * [Vocabulary](https://w3c.github.io/rdf-semantics/spec/#dfn-vocabulary) — A vocabulary is a set of [names](https://w3c.github.io/rdf-semantics/spec/#dfn-name).
+
+## New terms:
+
+* [Opaque semantics](/docs/semantics.md#opaque-semantics) – Any semantics for quoted triples where the syntactic form of the arguments to quoted triples is important.
+
+* [Semi-transparent semantics](/docs/semantics.md#semi-transparent-semantics) – Any semantics for quoted triples where the syntactic form of some, but not all, kinds of arguments to quoted triples is important _usually excepting blank nodes so that only their meaning (and not their syntactic form) is important_.
+
+* [Transparent semantics](/docs/semantics.md#transparent-semantics) – Any semantics for quoted triples where only the meaning of the arguments to quoted triples is important.

docs/semantics.md

@@ -0,0 +1,96 @@
+# RDF-star Semantics Options
+
+There have been several different semantics proposals for quoted triples, some with multiple semantics.  The proposed semantics differ in technical details but also differ in the entailments they support.  The major differences between semantics can be summarized in whether they are sensitive to the precise syntax of quoted triples or only to the semantics of the components of quoted triples.   The first kind of semantics have been called opaque and the second kind transparent.  There is also a semi-opaque (or semi-transparent) semantics, as in the community report, where the syntax of IRIs and literals are important but only the meaning of blank nodes is.
+
+The difference between the semantics can be seen in the following examples.
+
+Note that, although we use the Turtle syntax for convenience, we interpret it in a subtly different way: each blank node label is interpreted as a specific and fixed element of the set of blank nodes defined by the RDF abstract syntax.
+
+In RDF (extended to include `owl:sameAs` as equality to provide a way of requiring that two different IRIs mean the same thing)
+```
+:b                            owl:sameAs :c .
+_:a                           :b         "4"^^xsd:int .
+```
+entails
+```
+_:f                           :b         "4"^^xsd:int .
+```
+(because blank nodes can be replaced by blank nodes without changing their meaning)
+
+and 
+```
+_:a                           :b         "4"^^xsd:integer .
+```
+(because `xsd:int` and `xsd:integer` mean the same datatype as per the RDF specification)
+
+and 
+```
+_:a                           :c         "4"^^xsd:int .
+```
+(because `:b`and `:c` mean the same thing as per the `owl:sameAs` declaration above).
+
+
+## Transparent Semantics
+
+In the _transparent_ semantics
+```
+:b                            owl:sameAs :c .
+<< _:a :b "4"^^xsd:int     >> :d         :e .
+```
+entails
+```
+<< _:f :b "4"^^xsd:int     >> :d         :e .
+```
+and 
+```
+<< _:a :b "4"^^xsd:integer >> :d         :e .
+```
+and
+```
+<< _:a :c "4"^^xsd:int     >> :d         :e .
+```
+
+## Opaque Semantics
+
+In the _opaque_ semantics
+```
+:b owl:sameAs :c .
+<< _:a :b "4"^^xsd:int     >> :d         :e .
+```
+does not entail
+```
+<< _:f :b "4"^^xsd:int     >> :d         :e .
+```
+or 
+```
+<< _:a :b "4"^^xsd:integer >> :d         :e .
+```
+or
+```
+<< _:a :c "4"^^xsd:int     >> :d         :e .
+```
+
+## Semi-transparent Semantics
+
+In the _semi-transparent_ (also known as _semi-opaque_) semantics, as proposed in [the RDF-star final community group report](https://www.w3.org/2021/12/rdf-star.html)
+```
+:b owl:sameAs :c .
+<< _:a :b "4"^^xsd:int     >> :d         :e .
+```
+does entail
+```
+<< _:f :b "4"^^xsd:int     >> :d         :e .
+```
+but does not entail
+```
+<< _:a :b "4"^^xsd:integer >> :d         :e .
+```
+or
+```
+<< _:a :c "4"^^xsd:int     >> :d         :e .
+```
+
+In this semantics the IRI `:b` inside the quoted triple denotes (i.e. refers to, means) something in the real world just like it does in standard/un-quoted RDF triples. As per the `owl:sameAs` statement that meaning can also be refered to by `:c`. However, this semantics aims to faithfully record that the IRI `:b` was used to refer to that meaning and therefore doesn't support entailments that e.g. entail the usage of `:c` from the usage of of `:b` although they mean the same thing.
+
+One could also create other semi-opaque semantics where either IRIs or literals were transparent.
+